Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M310076200 on January 6, 2004

J. Biol. Chem., Vol. 279, Issue 12, 11917-11925, March 19, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/12/11917    most recent
M310076200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Runko, A. P.
Right arrow Articles by Sagerström, C. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Runko, A. P.
Right arrow Articles by Sagerström, C. G.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Isolation of nlz2 and Characterization of Essential Domains in Nlz Family Proteins*

Alexander P. Runko and Charles G. Sagerström{ddagger}

From the Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, Massachusetts 01605

Received for publication, September 10, 2003 , and in revised form, December 19, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
In this study, we first cloned nlz2, a second zebrafish member of the nlz-related zinc-finger gene family. nlz2 was expressed together with nlz1 in a broad posterior domain during gastrula stages as well as at the midbrain-hindbrain boundary and in the hindbrain caudal to rhombomere 4 during segmentation. nlz2 was also expressed in regions distinct from nlz1, notably in the forebrain, midbrain, and trunk. Misexpression of nlz2 in zebrafish embryos disrupted gene expression in the rostral hindbrain, similar to the effect of misexpressing nlz1. We next compared the nlz1 and nlz2 sequences to identify and characterize domains conserved within this family. We found a C-terminal domain required for nuclear localization and two conserved domains (the Sp motif and a putative C2H2 zinc finger) required for nlz1 function. We also demonstrate that Nlz1 self-associated via its C terminus, interacted with Nlz2, and bound to histone deacetylases. Last, we found two forms of Nlz1 generated from alternative translation initiation sites in vivo. These forms have distinct activities, apparently depending on the function of the N-terminal Sp motif. Our data demonstrate that nlz2 functions similarly to nlz1 and define conserved domains essential for nuclear localization, self-association, and corepressor binding in this novel family of zinc-finger genes.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
nlz (herein referred to as nlz1) was recently identified as a member of a novel subfamily of zinc-finger genes (1, 2). We have demonstrated that nlz1 functions to control segmental gene expression in the zebrafish hindbrain and likely represses transcription by recruiting the corepressor Groucho (3). Sequence similarities between members of this zinc-finger subfamily suggest that their functions are evolutionarily conserved and that the nlz-related genes nocA, elbow, and tlp-1 regulate embryonic development in Drosophila (4, 5) and Caenorhabditis elegans (6). Homozygous elbow and nocA mutants exhibit defects in tracheal development, particularly stalled and aberrant migration of the dorsal branch and lateral trunks of the trachea (5). Conversely, misexpression of elbow in the trachea results in the repression of genes expressed in the visceral branch and dorsal trunk, whereas the number of cells forming the dorsal branch increases (5). Homozygous nocA mutants also display brain abnormalities such as protrusion of the embryonic supraesophageal ganglion and a reduction of ocelli (4). Mutations in tlp-1 affect specification of asymmetric cell fates and cell fusion, resulting in abnormal tail morphogenesis in C. elegans (6). This family also includes two hypothetical proteins each in mouse and human, which have not yet been functionally characterized (1, 3, 5).

The Nlz subfamily of zinc-finger proteins is related to the vertebrate Sp1-like family of transcription factors (3, 6), and members of both families are expressed in developing embryos. However, although nlz-related genes have restricted expression patterns, sp1-related genes are more ubiquitously expressed. Sp1-like proteins function both as transcriptional activators and repressors, depending on the context (7); but to date, Nlz-related proteins have been implicated only in transcriptional repression (3, 5). Several domains are common to both families, including the Sp motif, the Btd (Buttonhead) box, and the C2H2 zinc-finger domain as well as serine/threonine- and glutamine-rich regions. Although some of these domains are essential for Sp1 function, their specific roles are not known, and it is not clear what role they may play in Nlz family proteins.

We recently reported the presence of a second gene in this family (nlz2) in the zebrafish expressed sequence tag data base (3). We have now cloned nlz2 and found high homology between nlz1 and nlz2. In particular, the Sp motif, the Btd box, and the putative C2H2 zinc finger are all conserved. We found that nlz2 was expressed together with nlz1 in a broad dorsoposterior domain in the zebrafish gastrula. During segmentation, nlz1 and nlz2 were coexpressed at the midbrain-hindbrain boundary, but the nlz2 expression pattern also diverged from that of nlz1. In particular, although nlz1 and nlz2 were both expressed in the caudal hindbrain, nlz2 expression never expanded rostrally to the rhombomere (r)1 5 segment, whereas nlz1 expression encompassed most of the hindbrain up to the r2 segment. Additionally, nlz2 was expressed in the forebrain, where nlz1 expression was never observed. We found that misexpression of nlz2 in zebrafish embryos resulted in loss of r3 gene expression in the hindbrain, similar to the effect of nlz1 misexpression (3). We next tested the function of domains conserved among members of the Nlz family. We found that nuclear localization of both Nlz2 and Nlz1 depended on an intact C terminus and that nuclear localization was required for optimal function. We found that Nlz1 bound Nlz2 and histone deacetylases (HDACs) and that Nlz1 self-associated, although self-association does not appear to be essential for function. Western blot analysis of endogenous Nlz1 revealed two forms, apparently generated by the use of alternative translation start sites. Further analyses determined that the two forms of Nlz1 have distinct activities in vivo and demonstrated that an N-terminal domain, shared with the Sp1 family of transcription factors, regulates Nlz1 function. Our results characterize a new member of the Nlz subfamily of zinc-finger proteins and define functionally important domains in this subfamily.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Reverse Transcription-PCR and Cloning—RNA was extracted from 10-h stage embryos using the ToTally RNA isolation kit (Ambion Inc.). cDNA was synthesized with Superscript II reverse transcriptase (Invitrogen) according to the manufacturer's instructions. The coding region of nlz2 was amplified with primers 5'-CCGCTCGAGAGATCTATGATCACATCGCCCTC-3' and 5'-TGCTCTAGACAATCACTGGTATCCAAGCG-3' from cDNA using the Expand High Fidelity PCR system (Roche Applied Science) according to the manufacturer's instructions. The amplified nlz2 product was purified using the QIAquick PCR purification kit (QIAGEN Inc.), digested with XhoI and XbaI, and subcloned into the same sites of pCS2+MT. The nlz2 construct was then digested with HindIII and XbaI, and the coding region (including the Myc epitope tags) was subcloned into the same sites of pCS2+MT-hsp70-NLS. The hsp70-nlz1 construct was generated by excising the nlz1 coding region (including the Myc epitope tags) with HindIII and NotI and subcloned by utilizing the same sites in pCS2+MT-hsp70-NLS. Nlz1 deletions constructs were generated as described previously (3) or by standard molecular biology cloning techniques into the pCS2+, pCS2+MT, or pCS2+MT-NLS vector. The N terminus of grg3 (nucleotides 1-795 of the open reading frame) was amplified with primers 5'-TAGAATTCAATGTATCCGCAGGGCCGGCAT-3' and 5'-GAAGGCCTCTCGTTGGACACATCCACCAC-3' from cDNA, digested with EcoRI and StuI, and subcloned into the same sites in pCS2+.

In Situ Hybridization and Immunostaining—Whole- and flat-mount in situ hybridizations using digoxigenin- or fluorescein-labeled antisense RNA probes were performed as described (3, 8), including a probe containing 1149 nucleotides of the nlz2 coding region (synthesized from expressed sequence tag fc59b07.y1, the sequence of which was cloned into the NotI and SalI sites of the pSPORT1 vector). Fluorescent immunostaining was performed as described (3, 9).

Embryo Injections—The synthesis and injection of capped mRNA encoding nlz deletion constructs were performed as described (3). Embryos were also injected with a DNA construct (hsp70-nlz2, hsp70-nlz1, or hsp70-green fluorescent protein (GFP), the coding region of which is under the control of a heat shock-inducible promoter) at the one- to two-cell stage, heat-shocked at 70% epiboly (~7.5 h post-fertilization (hpf)) for 1 h at 37 °C, and fixed at the 10-somite stage (14 hpf).

Glutathione S-Transferase (GST) Pull-down Assays and Western Analysis—GST-Nlz1 was constructed by excising the nlz1 coding region from pCS2+MT-nlz1 with EcoRI and NotI and subcloned into the pGEX-6P-3 vector (Amersham Biosciences) by utilizing the same restriction sites. The GST-Nlz1 fusion protein was expressed in BL21 cells (Amersham Biosciences) for 16 h at room temperature by the addition of 0.5 mM isopropyl-{beta}-D-thiogalactopyranoside. The bacteria were lysed, and the GST-Nlz1 protein was purified using a GST purification module (Amersham Biosciences) according to the manufacturer's instructions. Nlz1 constructs were labeled in vitro with [35S]methionine using the TNT SP6 coupled reticulocyte lysate system (Promega). GST pull-down assays were performed as described (3).

Anti-Nlz1 antiserum was derived from rabbits immunized with purified Nlz1 protein (Research Genetics). Western blot analyses confirmed that prebleed serum did not give any signal and that the anti-Nlz1 antiserum specifically recognized the Nlz1 protein. For Western blot analysis, lysates of three gastrula stage embryos or protein synthesized in vitro from the TNT SP6 coupled reticulocyte lysate system were used per lane. Samples were resolved on a 10% polyacrylamide gel by SDS-PAGE, Western-blotted, probed with a 1:5000 dilution of anti-Nlz antiserum and a 1:5000 dilution of horseradish peroxidase-linked donkey anti-rabbit Ig whole antibody (Amersham Biosciences), and detected using chemiluminescence.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Nlz2 Is Related to and Functions Similarly to Nlz1
Nlz2 and Nlz1 Share Conserved Domains—We have previously reported that an nlz1-related sequence, nlz2, can be assembled from a series of expressed sequence tags (3). Because none of these expressed sequence tags contain the entire nlz2 open reading frame, we cloned nlz2 from mid-gastrula stage zebrafish cDNA (see "Experimental Procedures"). Sequence analysis demonstrated that Nlz2 exhibits 55.3% identity to Nlz1 and that these proteins are distantly related to Sp1 family proteins. Nlz1, Nlz2, and Sp1 family proteins also share several conserved domains (Fig. 1). These include the Sp motif (a short N-terminal region found in the vertebrate Sp1-like family of transcription factors) (3, 6), the Btd box (an 8-11-amino acid motif that we initially identified as a potential zinc finger (3), but which more closely resembles a domain from Drosophila Buttonhead (10)), a putative C2H2 zinc finger (11-13), and serine/threonine-rich regions (14).



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 1.
Nlz1, Nlz2, and Sp1 proteins share conserved domains. Domains of the zebrafish Nlz1 (GenBankTM/EBI accession number AAK73547 [GenBank] and Nlz2 (accession number AY371081 [GenBank] ) proteins and the human Sp1 protein (accession number AAF67726 [GenBank] are shown schematically. The proteins are shown as white boxes, except for the internal start site (I; black), the Sp motif (Sp; gray), serine/threonine-rich regions (S/T; blue), the glutamine-rich region (Q; yellow), the Btd box (green), the C2H2 zinc finger (ZF; red), the repressor interaction domain (RID; hatched), and a region required for nuclear localization (N; stippled). Numbers indicate amino acid positions. Dr, Danio rerio; Hs, Homo sapiens.

 
nlz2 Is Coexpressed with nlz1 at the Gastrula Stages, but Expression Diverges by the Segmentation Stages—In situ hybridization analysis of zebrafish embryos revealed nlz2 mRNA early and throughout the blastomeres (Fig. 2, A and B), suggesting that it is maternally deposited. At early gastrula stages, nlz2 was coexpressed with nlz1 near the blastoderm margin with a gap in the dorsal midline (Fig. 2C) that was closed at mid-gastrula stages (Fig. 2D). At late gastrula stages, nlz2 was expressed in a broad posterior domain (Fig. 2E), similar to nlz1 (3). At early segmentation stages (11 hpf), nlz2 was expressed in the forebrain and in the posterior hindbrain caudal to r4 (Fig. 2, F and K), and expression extended throughout the whole length of the trunk (data not shown). By the three-somite stage (12 hpf), nlz2 expression was initiated at the midbrain-hindbrain boundary (MHB) (Fig. 2G). At this stage, expression of nlz2 in the forebrain also appeared to expand (Fig. 2, G and L). At the 10-somite stage (14 hpf), nlz2 expression was maintained in the forebrain, caudal hindbrain, and trunk, with expression at the MHB broadening anteriorly to encompass the midbrain (Fig. 2H). By the 14-somite stage (16 hpf), nlz2 expression in the midbrain was reduced, and expression in the forebrain extended to the optic primordium (Fig. 2I). At the 22-somite stage (20 hpf), nlz2 expression persisted in above-mentioned regions, including the optic stalk (Fig. 2J). In conclusion, we found that nlz2 was expressed similarly to nlz1 in the caudal hindbrain (although nlz2 never extended rostrally to r5) and at the MHB; however, it was also expressed in the forebrain, midbrain, and trunk.



View larger version (101K):
[in this window]
[in a new window]
  		FIG. 2.
nlz2 is dynamically expressed during zebrafish development. Embryos were processed for in situ hybridization for nlz2 (blue; A-L), nlz1 (red; C, D, and F-J), krox20 (red; K and L), and pax2.1 (red; K and L). Embryos were whole-mounted in A-E and flat-mounted in F-L and are shown in dorsal views with anterior to the top (magnification x20). Markers are shown in the lower left-hand corners, and embryonic stages are shown in the lower right-hand corners. fb, forebrain; mhb, midbrain-hindbrain boundary.

 
nlz2 Has Similar Activity to nlz1—We have reported that misexpression of nlz1 in zebrafish embryos disrupts gene expression in the rostral hindbrain (3). We next sought to determine whether nlz2 has a similar function. Because injecting nlz2 mRNA did not give high level expression of the Nlz2 protein, likely due to the mRNA being unstable, we instead misexpressed nlz2 under the control of a heat shock-inducible promoter (hsp70). Embryos were injected at the one- to two-cell stage with hsp70-nlz2, hsp70-nlz1, or hsp70-GFP; heat-shocked at 70% epiboly (~7.5 hpf) for 1 h at 37 °C; and fixed at the 10-somite stage (14 hpf). This led to efficient induction as shown by hsp70-GFP (Fig. 3A). We then assayed for defects in the rostral hindbrain by in situ analysis of krox20 expression (a marker for r3 and r5) (15). We found that misexpressing nlz1 under the control of the heat shock-inducible promoter repressed krox20 in r3, but not in r5 (Fig. 3C and Table I). This phenotype is identical to that observed upon injection of nlz1 mRNA (3), demonstrating that the heat shock-inducible construct works as well as mRNA injections. We found that misexpressing nlz2 resulted in a similar phenotype (Fig. 3D and Table I) and conclude that nlz2 has the same ability as nlz1 to disrupt gene expression in the rostral hindbrain.



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 3.
Misexpression of nlz2 disrupts r3 gene expression. Embryos were injected with 50 pg of hsp70-Myc-GFP (A and B), hsp70-Myc-nlz1 (C), or hsp70-Myc-nlz2 (D) DNA; heat-shocked at 70% epiboly for 1 h at 37 °C; and fixed at the 10-somite stage (14 hpf). Embryos were processed by in situ hybridization for krox20 (red; B-D), nlz1 (blue; C), and nlz2 (blue; D) expression or immunostained with anti-c-Myc antibody (A) to detect ectopic protein. Immunostaining confirmed wide distribution of induced protein (A). Ectopic nlz1 (C) and nlz2 (D) mRNAs were detected in injected embryos (low endogenous levels were undetectable). Markers or signals visualized are shown in the lower left-hand corners, and injected DNAs are shown in the lower right-hand corners. Arrows indicate loss of r3 gene expression. FITC, fluorescein isothiocyanate.

 


View this table:
[in this window]
[in a new window]
  		TABLE I
Misexpression of nlz2 disrupts r3 gene expression

One- to two-cell stage embryos were injected with 50 pg of hsp70-GFP, hsp70-nlz1, or hsp70-nlz2 DNA; heat-shocked at 70% epiboly for 1 h at 37 °C; fixed at the 10-somite stage; and analyzed by in situ hybridization for krox20 expression.

 
In Vivo Structure-Function Analysis Reveals Domains Essential for Nlz Function
The alignment in Fig. 1 defines several domains conserved between Nlz1 and Nlz2, and we next set out to define the functional importance of these domains. We elected to perform much of our structure-function analysis in zebrafish embryos because they provide a physiologically relevant environment and because functional analysis can be done rapidly (within 14 h after mRNA injection).

Nlz1 Function Requires the C2H2 Finger and Nuclear Translocation via a C-terminal Domain—Nlz1 is a nuclear protein (3), but lacks a consensus nuclear localization signal (NLS); we therefore set out to define the domain essential for nuclear localization. To this end, we determined the subcellular distribution of a panel of Nlz1 deletion constructs (summarized in Fig. 4). Embryos were injected with 800 pg of mRNA encoding Myc epitope-tagged Nlz1 deletion constructs at the one- to two-cell stage, flat-mounted at the early gastrula stage (~5 hpf), and examined for distribution of the Myc-tagged protein by fluorescent immunohistochemistry. We found that Nlz1{Delta}92C (data not shown) and Nlz1{Delta}173C (Fig. 5B), both of which contain C-terminal deletions, resided primarily in the cytoplasm instead of the nucleus, whereas N-terminal or internal deletions (e.g. Nlz1{Delta}60N, Nlz1{Delta}115N, Nlz1{Delta}116-234, Nlz1{Delta}238-384, and Nlz1{Delta}116-384) did not affect nuclear localization (summarized in Fig. 4) (data not shown). Specifically, the fact that Nlz1{Delta}92C is cytoplasmic, but Nlz1{Delta}23C is nuclear, indicates that amino acids 498-566 are important for nuclear localization of Nlz1. Because Nlz2 shows high homology to Nlz1, we reasoned that the Nlz2 C terminus might also govern nuclear localization. We found that Nlz2 is indeed a nuclear protein (Fig. 5C) and that nuclear localization is dependent on an intact C terminus because an Nlz2{Delta}180C construct (which lacks the C-terminal 180 amino acids) (data not shown) was primarily cytoplasmic (Fig. 5D). Several stretches of 4-7 amino acids are conserved between Nlz1 and Nlz2 in this region, but these domains do not correspond to motifs known to regulate nuclear localization (data not shown).



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 4.
Structure-function analysis reveals domains essential for Nlz1 function. Nlz1 deletion constructs were generated as described under "Experimental Procedures." Thick black lines indicate the Nlz1 protein, and thin gray lines indicate deleted amino acid regions. I (black), internal start site; Sp (gray), the Sp motif; S/T (blue), serine/threonine-rich regions; Q (yellow), the glutamine-rich region; Btd (green), the Btd box; ZF (red), the C2H2 zinc finger; RID (hatched), the repressor interaction domain; N (stippled), a region required for nuclear localization. Numbers indicate amino acid positions. The columns to the right indicate the percent of embryos exhibiting loss of r3 gene expression, the percent of embryos exhibiting overlap of r5- and r4-specific gene expression, cellular localization (primarily nuclear (N) or cytoplasmic (C)), self-association, and HDAC binding. The number of injected embryos examined for the in situ analysis (first and second columns) was >170 to <1275 for each construct (and >50 for the localization studies in the third column). The activities for Nlz1 and the Nlz1{Delta}385-460 and Nlz1{Delta}408-460 constructs were reported previously (3). +, strong association; +/-, weak interaction;-, very weak binding; ND, not determined. The asterisks denote the NLS sequence fused to the N terminus of Nlz1{Delta}92C and Nlz1{Delta}173C. All constructs contain six N-terminal Myc tags (data not shown). We have previously demonstrated that the constructs shown are expressed at comparable levels upon injection of mRNA into zebrafish embryos (3).

 



View larger version (70K):
[in this window]
[in a new window]
  		FIG. 5.
Nlz1 and Nlz2 are nuclear proteins whose localization depends on their respective C termini. Embryos were injected with 800 pg of Myc-nlz1 (A), Myc-nlz1{Delta}173C (B), Myc-nlz2 (C), or Myc-nlz2{Delta}180C (D) mRNA; fixed at 5 hpf; immunostained with anti-c-Myc antibody; and flat-mounted. Embryos are shown in animal pole views (magnification x40). A and C, Nlz1 and Nlz2 localize to the nucleus; B and D, Nlz1{Delta}173C and Nlz2{Delta}180C localize primarily to the cytoplasm. Injected RNAs are shown in the lower right-hand corners.

 
To determine whether Nlz1 function is dependent on nuclear localization, we assayed the function of constructs that localized to the cytoplasm (summarized in Fig. 4). Embryos were injected at the one- to two-cell stage with 400 pg of mRNA encoding each construct, fixed at the 10-somite stage (14 hpf), and assayed for loss of krox20 expression in r3 (same phenotype as shown in Fig. 3, C and D). We found that Nlz1{Delta}92C had reduced activity (19.7% of injected embryos affected versus 64.9% for wild-type Nlz1) (Fig. 4). However, upon addition of an NLS to Nlz1{Delta}92C, the frequency of affected embryos became comparable to that obtained with full-length Nlz1 (79.4% versus 64.9%, respectively) (Fig. 4). This suggests that Nlz1 must reside in the nucleus to be fully active. We also found that Nlz1{Delta}173C was likewise inactive (6.4%) (Fig. 4) and localized primarily to the cytoplasm (Fig. 5B). Interestingly, Nlz1{Delta}173C remained inactive even upon addition of an NLS (<5%) (Fig. 4), suggesting that amino acids 417-497 are necessary for Nlz1 function. Accordingly, a construct lacking the C2H2 zinc finger (amino acids 461-492, Nlz1{Delta}ZF) was less active than full-length Nlz1 (19.7% versus 64.9%, respectively) (Fig. 4), although it localized to the nucleus (Fig. 4) (data not shown). We conclude that the C terminus is required both for nuclear localization (via amino acids 498-566) and for function (via the C2H2 zinc finger).

nlz1 Is Coexpressed with a groucho-related Gene and Associates with Class I Histone Deacetylases—We have demonstrated that Nlz1 associates with the Groucho corepressor in vitro (3). We next used a probe for grg3 (the only zebrafish member of the groucho family isolated to date) (16, 17) to test whether groucho-related genes are coexpressed with nlz1 in vivo. We found that grg3 was expressed in the anterior region of the embryo, encompassing the forebrain and midbrain and extending into the rostral hindbrain (Fig. 6, A and B). This suggests that Grg3 might interact with Nlz1 in the rostral hindbrain and also raises the possibility that Nlz1 interacts with other corepressors farther caudally.



View larger version (59K):
[in this window]
[in a new window]
  		FIG. 6.
nlz1 is coexpressed with groucho and associates with class I histone deacetylases. A and B, nlz1 is coexpressed with groucho in the rostral hindbrain. Embryos were processed for double in situ hybridization for groucho (blue; A and B) together with nlz1 (red; A) or krox20 and pax2.1 (red; B). Embryos were flat-mounted and are shown in dorsal views with anterior to the top (magnification x20). Markers are shown in the lower left-hand corners, and the embryonic stage is shown in the lower right-hand corners. fb, forebrain; mb, midbrain; mhb, midbrain-hindbrain boundary. C, Nlz1 associates with class I histone deacetylases. Purified GST-HDAC2 fusion protein or GST was immobilized on glutathione-Sepharose beads and incubated with 35S-labeled Nlz1 or Nlz1 deletion constructs. After washing of the beads, bound proteins were separated by SDS-PAGE and visualized by autoradiography. Nlz interacted strongly with the GST-HDAC2 fusion protein, whereas Nlz1{Delta}173C, Nlz1{Delta}385-460, and Nlz1{Delta}408-460 exhibited weak binding. GST alone did not significantly bind to the Nlz1 deletion constructs. Aliquots containing 10% of the total amount of 35S-labeled protein used in the assay are shown (10% load).

 
We next tested whether Nlz1 binds HDACs, corepressors that function by removing acetyl groups from core histones to maintain a transcriptionally silenced chromatin state (18, 19). Utilizing GST pull-down assays, we found that Nlz1 associated with HDAC1 (data not shown) and HDAC2 (Fig. 6C), two class I HDACs. To determine the region of Nlz1 required for this interaction, we utilized our panel of Nlz1 deletion constructs and tested them for HDAC2 binding in the GST pull-down assay (Fig. 6C). We found that Nlz1 constructs lacking the C-terminal 173 amino acids exhibited significantly weaker binding to HDAC2 (Fig. 6C, lane 2). We also found that deletions within the C-terminal 206 amino acids (Fig. 6C, lanes 4-7) or the internal region (amino acids 116-384) (data not shown) did not affect HDAC2 binding. Instead, deletions of the region between the Btd box and the C2H2 zinc finger (Fig. 6C, lanes 8 and 9) severely affected HDAC2 binding. We note that this region is also involved in Groucho binding (3); and because Groucho mediates repression by recruiting HDACs (18, 19), this might point to a synergistic role for Groucho and HDACs in regulating Nlz1 function.

Nlz1 Self-associates and Binds to Nlz2—Because the Nlz-related protein Elbow was found to self-associate (5), we used GST pull-down assays to test whether Nlz1 self-associates (Fig. 7A). We found that Nlz1 self-associated efficiently (Fig. 7A, lane 1). To define the domain required for association, we utilized our panel of Nlz1 deletion constructs and tested them in the GST pull-down assay (Fig. 7A, lanes 2-10). We found that Nlz1 self-association was dependent on an intact C terminus. In particular, deletions within the N terminus had no effect on association (Fig. 7A, lanes 2 and 3), whereas deletions within the C terminus exhibited weaker binding (lanes 4-6, 9, and 10), except for Nlz1{Delta}23C (data not shown). Deletion of either the Btd box or the C2H2 zinc finger alone did not significantly decrease Nlz1 interaction (Fig. 7A, lanes 7 and 8). Taken together, these results suggest that several regions required for self-association are located within the C-terminal half of Nlz1.



View larger version (21K):
[in this window]
[in a new window]
  		FIG. 7.
Nlz1 self-associates and binds to Nlz2. A, Nlz1 self-association requires an intact C terminus. Purified GST-Nlz1 fusion protein or GST was immobilized on glutathione-Sepharose beads and incubated with 35S-labeled Nlz1or Nlz1 deletion constructs. After washing of the beads, bound proteins were separated by SDS-PAGE and visualized by autoradiography. Nlz1 interacted strongly with the GST-Nlz1 fusion protein, whereas Nlz1{Delta}238-384, Nlz1{Delta}385-460, Nlz1{Delta}408-460, Nlz1{Delta}92C, and Nlz1{Delta}173C exhibited weak binding. GST alone did not significantly bind to the Nlz1 deletion constructs. Aliquots containing 10% of the total amount of 35S-labeled protein used in the assay are shown (10% load). B, Nlz2 interacts with Nlz1. Nlz2 interacted strongly with the GST-Nlz1 fusion protein and did not significantly bind to GST alone.

 
To determine whether self-association is important for Nlz1 function, we misexpressed Nlz1 deletion constructs exhibiting weaker association (Nlz1{Delta}238-384, Nlz1{Delta}385-460, Nlz1{Delta}408-460, Nlz1{Delta}92C, and Nlz1{Delta}173C) and assayed their ability to disrupt krox20 expression in r3 of the hindbrain (same phenotype as in Fig. 3, C and D; data summarized in Fig. 4). Some of these constructs were as active as full-length Nlz1 (e.g. Nlz1{Delta}238-384, 41.6%) (Fig. 4), suggesting that self-association is not essential for Nlz1 function. Although some deletion constructs displayed a pronounced reduction in activity (e.g. Nlz1{Delta}408-460, 25.7%; Nlz1{Delta}92C, 19.7%; Nlz1{Delta}173C, 6.4%; and Nlz1{Delta}385-460, <5%), this is likely explained by the deletion of other essential domains (such as the C2H2 zinc finger) in those constructs. Thus, we did not found a correlation between self-association and function, suggesting that self-association is not necessary for Nlz1 function under these conditions.

We also examined whether Nlz1 interacts with Nlz2 utilizing GST pull-down assays and found that Nlz2 and Nlz1 did form complexes (Fig. 7B, lane 1). Because Elbow and NocA are thought to function as a complex in repressing transcription (5), an Nlz1-Nlz2 complex may function in the caudal hindbrain or at the MHB, where nlz1 and nlz2 are coexpressed.

Two forms of Nlz1 with Distinct N Termini and Disparate Activities Are Present in Vivo
Two Forms of Nlz1 Are Present in Vivo—Western analysis using polyclonal rabbit antiserum raised against full-length Nlz1 revealed two major bands in lysates from gastrula stage embryos (Fig. 8A, lane 1): a 72-kDa form close to the predicted size for Nlz1 and a lower molecular mass form (62 kDa). We found these two forms in embryo lysates from the 50% epiboly stage (5 hpf) to at least the 10-somite stage (14 hpf) (data not shown). We also found that in vitro translation of a plasmid containing the nlz1 cDNA yielded the same two forms, but with the smaller form predominating (Fig. 8A, lane 2). We conclude that two forms of Nlz1 are present in vivo and can be generated from the same cDNA.



View larger version (46K):
[in this window]
[in a new window]
  		FIG. 8.
Nlz1 exists in two forms with distinct activities. A, two forms of Nlz1 are detected in vivo and in vitro. Lysates of three uninjected embryos (lane 1), 1 µl of in vitro translated Nlz1 (lane 2), and 1 µl of a mock in vitro translation reaction (lane 3) were resolved on a 10% polyacrylamide gel by SDS-PAGE; Western-blotted; probed with anti-Nlz1 antiserum; and detected using chemiluminescence. The anti-Nlz1 antiserum did not cross-react with Nlz2 (data not shown). B, the short form of Nlz1 is generated from a downstream initiation site at position 61 of Nlz1. 5 µl of in vitro translated 35S-labeled Nlz1 (lane 1), Nlz1Met->Gly-Pro (lane 2), Myc-Nlz1 (lane 3), and Myc-Nlz1{Delta}60N (lane 4) were separated by SDS-PAGE and visualized by autoradiography. C-G, misexpression of Nlz1 constructs with N-terminal deletions disrupts gene expression at the r4/r5 boundary. Embryos were injected with 400 pg of {beta}-galactosidase mRNA ({beta}gal; C) or mRNA encoding Nlz1{Delta}60N (D and E), Nlz1{Delta}385-460 (F), or Nlz1{Delta}88-115 (G). Injected embryos were fixed at 14 hpf and processed by in situ hybridization for krox20 (expressed in r3 and r5; red) and hoxb1a (expressed in r4; blue). Embryos were flat-mounted and are shown in dorsal views with the anterior to the top. White arrows point to overlap in hoxb1a and krox20 expression at the r4/r5 boundary. Black arrows point to loss of r3 gene expression.

 
Alternative Translation Initiation Sites Generate the Two Forms of Nlz1—We reasoned that the two forms of Nlz1 might be the result of translation initiating from two distinct start sites in the nlz1 transcript, perhaps due to "leaky scanning" (20, 21). In support of this idea, a downstream AUG codon is present at residue 61, and translation from this site is predicted to generate a 62-kDa protein. To test this possibility, we mutated the methionine at position 61 to glycine and proline (nlz1Met->Gly-Pro). We found that nlz1Met->Gly-Pro yielded the full-length Nlz1 protein, but not the smaller 62-kDa form (Fig. 8B, lane 2), consistent with initiation at the downstream start site being eliminated. In this lane, a smaller Nlz1 form was present, perhaps due to initiation at a site even farther downstream. We conclude that two forms of Nlz1 are generated by translation from two initiation sites in the nlz1 transcript.

We noted that in vitro translation of Myc-tagged Nlz1 (Fig. 4) yielded a single protein of 90 kDa (Fig. 8B, lane 3). This corresponds closely to the size of full-length Nlz1 plus the Myc tags, suggesting that introduction of Myc tags at the N terminus prevents generation of the smaller form of Nlz1. This is likely due to the initiation codon contributed by the Myc tags being more similar to the consensus sequence for initiation codons (Kozak sequence) (22-24) than the translation start sites present in endogenous nlz1 and the nlz1 plasmid (data not shown).

The Two Forms of Nlz1 Have Distinct Activities—Because we detect two forms of Nlz1 in the embryo, we next examined whether the two forms have the same activity. To this end, we used the Nlz1{Delta}60N construct (Fig. 4), which lacks the first 60 amino acids and encodes only the small form of Nlz1 (Fig. 8B, lane 4), and compared it with Myc-Nlz1, which generates only the full-length form (Fig. 8B, lane 3). We microinjected 400 pg of mRNA encoding either construct into one- to two-cell stage zebrafish embryos and assayed for defects in the hindbrain. We found that Nlz1{Delta}60N disrupted r3 gene expression (49.3% of embryos) (Fig. 8, D and E), similar to full-length Nlz1 (Fig. 3C; summarized in Fig. 4). However, about half (46.6%) of the nlz1{Delta}60N-injected embryos (Fig. 8, D and E) also exhibited overlap in r4- and r5-specific gene expression, with occasional lateral expansion of gene expression. We have previously demonstrated that expression of Nlz1 constructs lacking the corepressor-binding site or of an Nlz1 construct fused to the VP16 activation domain also gives rise to this phenotype (Fig. 8F) (3). VP16-Nlz1 likely activates transcription of Nlz1-regulated genes directly, whereas Nlz1 constructs lacking the corepressor-binding site appear to do so indirectly, by inhibiting repression mediated by wild-type Nlz (3). Thus, the short form of Nlz1 retains the ability to disrupt gene expression in r3, but may also promote expression of Nlz1-regulated genes under some circumstances.

An N-terminal Motif Is Important for Nlz1 Function—Although the short form of Nlz1 (Nlz1{Delta}60N) induced the same phenotype as constructs that promoted expression of Nlz1-regulated genes, its ability to bind corepressors was unaffected (Fig. 4). This suggests that the N terminus of Nlz1 might contain a domain required for Nlz1 to repress transcription. To define such a domain, we misexpressed constructs with N-terminal deletions in zebrafish embryos and assayed for changes in hindbrain gene expression (summarized in Fig. 4). We found that removal of the Sp motif was sufficient to promote overlap in r4/r5 gene expression (37.7% of nlz1{Delta}88-115-injected embryos affected) (Fig. 8G), suggesting that this domain is required for Nlz1-mediated repression.

We noted that truncations that removed the Sp motif (nlz1{Delta}88-115 and nlz1{Delta}115N) gave rise primarily to overlap in r4/r5 gene expression (37.7 and 41.9%, respectively) (Fig. 4). In contrast, truncations that removed domains close to the Sp motif (nlz1{Delta}60N and nlz1{Delta}116-234) not only promoted overlap in r4/r5 gene expression (46.6 and 46.2%, respectively) (Fig. 4), but also disrupted r3 gene expression (49.3 and 32.2%, respectively) (Fig. 4). One exception may be Nlz1{Delta}116-384, but this construct represents a large deletion (~50% of the protein). We suggest that removal of adjacent sequences may partially interfere with the Sp motif to generate forms of Nlz1 that have both functions (albeit with reduced activity). Notably, the short form of Nlz1 is present in vivo and may modulate the expression of Nlz1-regulated genes in the embryo.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
In this study, we have shown that nlz2 and nlz1 share conserved domains, are coexpressed during hindbrain development, and are both able to disrupt gene expression in the rostral hindbrain upon misexpression in zebrafish embryos. We used structure-function analysis to determine that nuclear localization of Nlz2 and Nlz1 is governed by their respective C termini and that Nlz1 function is enhanced when residing in the nucleus. We found that Nlz1 associates with itself as well as with Nlz2, although self-association does not appear to be essential for tested Nlz1 activities. We have previously shown that Nlz1 binds Groucho (3), but here we found that nlz1 was not always coexpressed with a groucho-related gene (grg3) and instead may associate directly with histone deacetylases in some regions of the embryo. Two forms of Nlz1 resulting from translation initiating at two distinct start sites in the nlz1 transcript were detected in vivo. These forms possess distinct functions as a result of differential activity of the Sp1 domain, an N-terminal domain required for Nlz1 function. We conclude that nlz genes require several conserved domains to regulate development of the zebrafish hindbrain.

Nlz Family Zinc-finger Proteins Likely Act as Transcriptional Repressors—We found that misexpression of nlz1 or nlz2 in zebrafish embryos resulted in loss of genes expressed in r2 and r3 of the hindbrain (Fig. 3, C and D) (3). Conversely, disruption of nlz function led to expansion of r5 and, to a lesser extent, r3 gene expression into r4 (Fig. 8, C-G) (3). Similarly, nocA and elbow mutant flies display expanded expression of trachea-specific genes (4, 5). Thus, the biological activities of elbow and nocA in the Drosophila trachea and of nlz1 and nlz2 in the zebrafish hindbrain are consistent with a role in repression of transcription, although it is not clear if this will hold true for all nlz subfamily genes because tlp-1 in C. elegans and nlz-related genes in mouse and human have not yet been studied extensively.

Nlz family proteins also associate with corepressors. Elbow associates with Groucho (5), and both Nlz1 and Nlz2 bind Groucho as well as class I histone deacetylases (Fig. 6C) (3). These interactions appear to be essential because Nlz1 constructs unable to bind corepressors lack the activity of wild-type Nlz1 and instead display dominant-negative activity (3). Notably, a groucho-related gene (grg3) was expressed in the zebrafish rostral hindbrain (Fig. 6, A and B), whereas HDAC1 was expressed ubiquitously,2 suggesting that Nlz1 may bind different corepressors in different regions of the embryo. Finally, repressors are expected to function in the nucleus; and consistent with this, Elbow (5), TLP-1 (6), Nlz1 (3), and Nlz2 (Fig. 5C) localize to the nucleus. Additionally, we found that optimal Nlz1 function was dependent on nuclear localization. Collectively, these results suggest that Nlz-related proteins repress transcription by recruiting corepressors.

We found that the putative C2H2 zinc finger is important for Nlz1 function (Fig. 4), but it is not clear if Nlz-related proteins bind to DNA because they contain only one such zinc finger, whereas known DNA-binding proteins usually contain several (12). The association of several Nlz-related proteins might provide a sufficient number of zinc fingers for DNA binding, and we found that Nlz1 self-associated (Fig. 7A) and also interacted with Nlz2 (Fig. 7B). Similarly, Drosophila Elbow self-associates and binds to NocA (5). However, our experiments suggest that self-association is not necessary for Nlz function (Fig. 4), and the C2H2 finger of the Nlz-related proteins displays differences relative to the consensus for DNA-binding C2H2 fingers. For instance, amino acids at positions -1, 3, and 6 of the consensus contact the primary DNA strand, and the amino acid at position 2 contacts the complementary strand (25), but only positions 2 and 3 are conserved in the Nlz family. This suggests that the C2H2 zinc finger of Nlz-related proteins does not bind DNA, although we cannot exclude the possibility that residues flanking the zinc-finger domain might be involved in DNA recognition (26, 27). Instead, we favor the possibility that the single zinc finger of Nlz-related proteins mediates protein-protein interactions, as reported for other C2H2 zinc-finger proteins (28-30), and that Nlz proteins are recruited to target genes by other DNA-bound factors.

Nlz Proteins Are Broadly Related to the Sp1 Family of Transcription Factors—Sequence analysis suggests that nlz1 and nlz2 are distantly related to the Sp1 family of transcription factors. The vertebrate Sp1 family encodes transcription factors containing three C2H2 fingers (31), and mouse knockouts have demonstrated that these genes are important for embryonic development (32-35). For instance, sp1 knockout mice display retarded growth and severe abnormalities during development, correlating with the fact that Sp1 activates transcription of numerous genes (36). Furthermore, the sp1-related gene sp4 is expressed in the central nervous system, and mice lacking the conserved zinc-finger domains of Sp4 exhibit low postnatal survival rates, impaired growth, and fertility defects (34).

Interestingly, although Nlz-related proteins likely function as repressors, most Sp1-like proteins mediate activation of transcription. One exception is Sp3, which activates transcription in an in vitro assay (10, 37), but also contains an inhibitory domain (amino acid triplet KEE) that silences its own glutamine-rich activation domain (37). Other vertebrate sp1-like genes include zebrafish bts1 (buttonhead/sp-related-1), which is involved in induction and early maintenance of pax2.1 expression at the MHB (38), and two Drosophila Sp1-related proteins (Buttonhead and D-Sp1) (39-41). However, these genes do not contain regions homologous to the Sp motif and are more distantly related to the Nlz and Sp1 families (6). XSPR-1, a novel Sp1-related protein in Xenopus laevis, is expressed in the neuroectoderm (as well as in the forebrain and otic vesicles and at the MHB) during embryogenesis (42).

The vertebrate Sp1 family shares several conserved domains with the Nlz subfamily. First, Sp1 proteins contain three C2H2 zinc fingers, which bind GC/GT boxes widely distributed in promoters and enhancers (43-47). Although Nlz-related proteins contain only one putative C2H2 zinc finger, which may mediate protein-protein interactions, this domain is required for Nlz1 function (Fig. 4). Second, the Btd box is present in all vertebrate Sp1-like proteins. Nlz family proteins contain a domain that we originally identified as a putative zinc finger (3), but that actually resembles the Btd box. Deletion of this domain in Nlz1 (together with an adjacent region) reduces its affinity for the corepressors Groucho (3) and HDACs (Fig. 6C) and instead elicits a dominant-negative effect (3). Interestingly, deletion of a small region (24 or 31 amino acids, including the Btd box) immediately adjacent to the three zinc fingers in Sp1 results in a reduction of transcriptional activation (10, 48). Third, Sp1-like factors have multiple serine/threonine- and glutamine-rich regions, which mediate transcriptional activation (14, 49). Although Nlz1 also exhibits multiple serine/threonine-rich regions (amino acids 112-137, 181-201, 244-289, and 425-445) and a glutamine-rich region (amino acids 203-227), we did not found a specific role for these domains (Fig. 4). Last, the Sp motif is shared between the Sp1 and Nlz families (3, 6). Although deletion of the Sp motif (or adjacent regions) creates a form of Nlz1 that may lack repressor activity, deletion of N-terminal regions including the Sp domain does not significantly affect the transcriptional activity of Sp1 (48). However, p74 specifically binds the N terminus of Sp1 and inhibits Sp1-mediated transcription in vivo (50), raising the possibility that Nlz1 function is also regulated by factors that bind its N terminus. Taken together, our observations indicate that Sp1 and Nlz family proteins utilize similar functional domains, although members of these families may have different effects on transcription (activation versus repression), perhaps as a result of interacting with different cofactors. Furthermore, some conserved domains may perform different functions in different proteins (e.g. the Sp1 zinc fingers bind DNA, whereas the Nlz zinc finger may bind proteins).


   FOOTNOTES
 
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EBI Data Bank with accession number(s) AY371081 [GenBank] .

* This work was supported by National Institutes of Health Grants NS38183 and HD39156 (to C. G. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Pharmacology, LRB 822, 364 Plantation St., Worcester, MA 01605. Tel.: 508-856-8006; Fax: 508-856-8007; E-mail: charles.sagerstrom{at}umassmed.edu.

1 The abbreviations used are: r, rhombomere; HDAC, histone deacetylase; GFP, green fluorescent protein; hpf, hours post-fertilization; GST, glutathione S-transferase; MHB, midbrain-hindbrain boundary; NLS, nuclear localization signal. Back

2 S. Zchut and C. G. Sagerström, unpublished data. Back


   ACKNOWLEDGMENTS
 
We thank members of the Sagerström laboratory for helpful discussions, Yasemin Gunes for assistance in cloning nlz2, and Scott Wolfe for discussions on zinc-finger proteins.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Andreazzoli, M., Broccoli, V., and Dawid, I. B. (2001) Mech. Dev. 104, 117-120[CrossRef][Medline] [Order article via Infotrieve]
  2. Sagerström, C. G., Kao, B. A., Lane, M. E., and Sive, H. (2001) Dev. Dyn. 220, 402-408[CrossRef][Medline] [Order article via Infotrieve]
  3. Runko, A. P., and Sagerström, C. G. (2003) Dev. Biol. 262, 254-267[CrossRef][Medline] [Order article via Infotrieve]
  4. Cheah, P. Y., Meng, Y. B., Yang, X., Kimbrell, D., Ashburner, M., and Chia, W. (1994) Mol. Cell. Biol. 14, 1487-1499[Abstract/Free Full Text]
  5. Dorfman, R., Glazer, L., Weihe, U., Wernet, M. F., and Shilo, B. Z. (2002) Development 129, 3585-3596[Abstract/Free Full Text]
  6. Zhao, X., Yang, Y., Fitch, D. H., and Herman, M. A. (2002) Development 129, 1497-1508[Medline] [Order article via Infotrieve]
  7. Philipsen, S., and Suske, G. (1999) Nucleic Acids Res. 27, 2991-3000[Abstract/Free Full Text]
  8. Sagerström, C. G., Grinblat, Y., and Sive, H. (1996) Development 122, 1873-1883[Abstract]
  9. Hatta, K. (1992) Neuron 9, 629-642[CrossRef][Medline] [Order article via Infotrieve]
  10. Athanikar, J. N., Sanchez, H. B., and Osborne, T. F. (1997) Mol. Cell. Biol. 17, 5193-5200[Abstract]
  11. Klug, A., and Rhodes, D. (1987) Cold Spring Harbor Symp. Quant. Biol. 52, 473-482[Abstract/Free Full Text]
  12. Berg, J. M. (1990) J. Biol. Chem. 265, 6513-6516[Free Full Text]
  13. Iuchi, S. (2001) Cell Mol. Life Sci. 58, 625-635[CrossRef][Medline] [Order article via Infotrieve]
  14. Pascal, E., and Tjian, R. (1991) Genes Dev. 5, 1646-1656[Abstract/Free Full Text]
  15. Oxtoby, E., and Jowett, T. (1993) Nucleic Acids Res. 21, 1087-1095[Abstract/Free Full Text]
  16. Kobayashi, M., Nishikawa, K., Suzuki, T., and Yamamoto, M. (2001) Dev. Biol. 232, 315-326[CrossRef][Medline] [Order article via Infotrieve]
  17. Wülbeck, C., and Campos-Ortega, J. A. (1997) Dev. Genes Evol. 207, 156-166[CrossRef]
  18. Chen, G., and Courey, A. J. (2000) Gene (Amst.) 249, 1-16[CrossRef][Medline] [Order article via Infotrieve]
  19. Courey, A. J., and Jia, S. (2001) Genes Dev. 15, 2786-2796[Free Full Text]
  20. Kozak, M. (1981) Nucleic Acids Res. 9, 5233-5262[Abstract/Free Full Text]
  21. Sedman, S. A., Gelembiuk, G. W., and Mertz, J. E. (1990) J. Virol. 64, 453-457[Abstract/Free Full Text]
  22. Kozak, M. (1987) Nucleic Acids Res. 15, 8125-8148[Abstract/Free Full Text]
  23. Kozak, M. (1987) J. Mol. Biol. 196, 947-950[CrossRef][Medline] [Order article via Infotrieve]
  24. Kozak, M. (1997) EMBO J. 16, 2482-2492[CrossRef][Medline] [Order article via Infotrieve]
  25. Wolfe, S. A., Nekludova, L., and Pabo, C. O. (2000) Annu. Rev. Biophys. Biomol. Struct. 29, 183-212[CrossRef][Medline] [Order article via Infotrieve]
  26. Pedone, P. V., Ghirlando, R., Clore, G. M., Gronenborn, A. M., Felsenfeld, G., and Omichinski, J. G. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 2822-2826[Abstract/Free Full Text]
  27. Omichinski, J. G., Pedone, P. V., Felsenfeld, G., Gronenborn, A. M., and Clore, G. M. (1997) Nat. Struct. Biol. 4, 122-132[CrossRef][Medline] [Order article via Infotrieve]
  28. Yang, H. Y., and Evans, T. (1992) Mol. Cell. Biol. 12, 4562-4570[Abstract/Free Full Text]
  29. Gregory, R. C., Taxman, D. J., Seshasayee, D., Kensinger, M. H., Bieker, J. J., and Wojchowski, D. M. (1996) Blood 87, 1793-1801[Abstract/Free Full Text]
  30. Merika, M., and Orkin, S. H. (1995) Mol. Cell. Biol. 15, 2437-2447[Abstract]
  31. Kadonaga, J. T., Carner, K. R., Masiarz, F. R., and Tjian, R. (1987) Cell 51, 1079-1090[CrossRef][Medline] [Order article via Infotrieve]
  32. Bouwman, P., Gollner, H., Elsasser, H. P., Eckhoff, G., Karis, A., Grosveld, F., Philipsen, S., and Suske, G. (2000) EMBO J. 19, 655-661[CrossRef][Medline] [Order article via Infotrieve]
  33. Gollner, H., Dani, C., Phillips, B., Philipsen, S., and Suske, G. (2001) Mech. Dev. 106, 77-83[CrossRef][Medline] [Order article via Infotrieve]
  34. Supp, D. M., Witte, D. P., Branford, W. W., Smith, E. P., and Potter, S. S. (1996) Dev. Biol. 176, 284-299[CrossRef][Medline] [Order article via Infotrieve]
  35. Harrison, S. M., Houzelstein, D., Dunwoodie, S. L., and Beddington, R. S. (2000) Dev. Biol. 227, 358-372[CrossRef][Medline] [Order article via Infotrieve]
  36. Marin, M., Karis, A., Visser, P., Grosveld, F., and Philipsen, S. (1997) Cell 89, 619-628[CrossRef][Medline] [Order article via Infotrieve]
  37. Dennig, J., Beato, M., and Suske, G. (1996) EMBO J. 15, 5659-5667[Medline] [Order article via Infotrieve]
  38. Tallafuss, A., Wilm, T. P., Crozatier, M., Pfeffer, P., Wassef, M., and Bally-Cuif, L. (2001) Development 128, 4021-4034[Medline] [Order article via Infotrieve]
  39. Wimmer, E. A., Frommer, G., Purnell, B. A., and Jackle, H. (1996) Mech. Dev. 59, 53-62[CrossRef][Medline] [Order article via Infotrieve]
  40. Wimmer, E. A., Jackle, H., Pfeifle, C., and Cohen, S. M. (1993) Nature 366, 690-694[CrossRef][Medline] [Order article via Infotrieve]
  41. Schock, F., Purnell, B. A., Wimmer, E. A., and Jackle, H. (1999) Mech. Dev. 89, 125-132[CrossRef][Medline] [Order article via Infotrieve]
  42. Ossipova, O., Stick, R., and Pieler, T. (2002) Mech. Dev. 115, 117-122[CrossRef][Medline] [Order article via Infotrieve]
  43. Dynan, W. S., and Tjian, R. (1983) Cell 35, 79-87[CrossRef][Medline] [Order article via Infotrieve]
  44. Dynan, W. S., and Tjian, R. (1983) Cell 32, 669-680[CrossRef][Medline] [Order article via Infotrieve]
  45. Gidoni, D., Dynan, W. S., and Tjian, R. (1984) Nature 312, 409-413[CrossRef][Medline] [Order article via Infotrieve]
  46. Gidoni, D., Kadonaga, J. T., Barrera-Saldana, H., Takahashi, K., Chambon, P., and Tjian, R. (1985) Science 230, 511-517[Abstract/Free Full Text]
  47. Thiesen, H. J., and Bach, C. (1990) Nucleic Acids Res. 18, 3203-3209[Abstract/Free Full Text]
  48. Courey, A. J., and Tjian, R. (1988) Cell 55, 887-898[CrossRef][Medline] [Order article via Infotrieve]
  49. Courey, A. J., Holtzman, D. A., Jackson, S. P., and Tjian, R. (1989) Cell 59, 827-836[CrossRef][Medline] [Order article via Infotrieve]
  50. Murata, Y., Kim, H. G., Rogers, K. T., Udvadia, A. J., and Horowitz, J. M. (1994) J. Biol. Chem. 269, 20674-20681[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Proc. Natl. Acad. Sci. USAHome page
J. D. Brown, S. Dutta, K. Bharti, R. F. Bonner, P. J. Munson, I. B. Dawid, A. L. Akhtar, I. F. Onojafe, R. P. Alur, J. M. Gross, et al.
Expression profiling during ocular development identifies 2 Nlz genes with a critical role in optic fissure closure
PNAS, February 3, 2009; 106(5): 1462 - 1467.
[Abstract] [Full Text] [PDF]

Home page
 DevelopmentHome page
S.-J. Ji, G. Periz, and S. Sockanathan
Nolz1 is induced by retinoid signals and controls motoneuron subtype identity through distinct repressor activities
Development, January 15, 2009; 136(2): 231 - 240.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/12/11917    most recent
M310076200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Runko, A. P.
Right arrow Articles by Sagerström, C. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Runko, A. P.
Right arrow Articles by Sagerström, C. G.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2004 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement