The topoisomerase II inhibitor VM-26 induces marked changes in histone H1 kinase activity, histones H1 and H3 phosphorylation, and chromosome condensation in G2 phase and mitotic BHK cells.

We have examined the effects of topoisomerase inhibitors on the phosphorylation of histones in chromatin during the G2 and the M phases of the cell cycle. Throughout the G2 phase of BHK cells, addition of the topoisomerase II inhibitor VM-26 prevented histone H1 phosphorylation, accompanied by the inhibition of intracellular histone H1 kinase activity. However, VM-26 had no inhibitory effect on the activity of the kinase in vitro, suggesting an indirect influence on histone H1 kinase activity. Entry into mitosis was also prevented, as monitored by the absence of nuclear lamina depolymerization, chromosome condensation, and histone H3 phosphorylation. In contrast, the topoisomerase I inhibitor, camptothecin, inhibited histone H1 phosphorylation and entry into mitosis only when applied at early G2. In cells that were arrested in mitosis, VM-26 induced dephosphorylation of histones H1 and H3, DNA breaks, and partial chromosome decondensation. These changes in chromatin parameters probably reverse the process of chromosome condensation, unfolding condensed regions to permit the repair of strand breaks in the DNA that were induced by VM-26. The involvement of growth-associated histone H1 kinase in these processes raises the possibility that the cell detects breaks in the DNA through their effects on the state of DNA supercoiling in constrained domains or loops. It would appear that histone H1 kinase and topoisomerase II work coordinately in both chromosome condensation and decondensation, and that this process participates in the VM-26-induced G2 arrest of the cell.

Human sperm telomere-binding complex involves histone H2B and secures telomere membrane attachment.

Telomeres are unique chromatin domains located at the ends of eukaryotic chromosomes. Telomere functions in somatic cells involve complexes between telomere proteins and TTAGGG DNA repeats. During the differentiation of germ-line cells, telomeres undergo significant reorganization most likely required for additional specific functions in meiosis and fertilization. A telomere-binding protein complex from human sperm (hSTBP) has been isolated by detergent treatment and was partially purified. hSTBP specifically binds double-stranded telomeric DNA and does not contain known somatic telomere proteins TRF1, TRF2, and Ku. Surprisingly, the essential component of this complex has been identified as a specific variant of histone H2B. Indirect immunofluorescence shows punctate localization of H2B in sperm nuclei, which in part coincides with telomeric DNA localization established by fluorescent in situ hybridization. Anti-H2B antibodies block interactions of hSTBP with telomere DNA, and spH2B forms specific complex with this DNA in vitro, indicating that this protein plays a role in telomere DNA recognition. We propose that hSTBP participates in the membrane attachment of telomeres that may be important for ordered chromosome withdrawal after fertilization.

Requirement for p34cdc2 kinase is restricted to mitosis in the mammalian cdc2 mutant FT210.

The mouse FT210 cell line is a temperature-sensitive cdc2 mutant. FT210 cells are found to arrest specifically in G2 phase and unlike many alleles of cdc2 and cdc28 mutants of yeasts, loss of p34cdc2 at the nonpermissive temperature has no apparent effect on cell cycle progression through the G1 and S phases of the division cycle. FT210 cells and the parent wild-type FM3A cell line each possess at least three distinct histone H1 kinases. H1 kinase activities in chromatography fractions were identified using a synthetic peptide substrate containing the consensus phosphorylation site of histone H1 and the kinase subunit compositions were determined immunochemically with antisera prepared against the "PSTAIR" peptide, the COOH-terminus of mammalian p34cdc2 and the human cyclins A and B1. The results show that p34cdc2 forms two separate complexes with cyclin A and with cyclin B1, both of which exhibit thermal lability at the non-permissive temperature in vitro and in vivo. A third H1 kinase with stable activity at the nonpermissive temperature is comprised of cyclin A and a cdc2-like 34-kD subunit, which is immunoreactive with anti-"PSTAIR" antiserum but is not recognized with antiserum specific for the COOH-terminus of p34cdc2. The cyclin A-associated kinases are active during S and G2 phases and earlier in the division cycle than the p34cdc2-cyclin B1 kinase. We show that mouse cells possess at least two cdc2-related gene products which form cell cycle regulated histone H1 kinases and we propose that the murine homolog of yeast p34cdc/CDC28 is essential only during the G2-to-M transition in FT210 cells.

Sequence-specific packaging of DNA in human sperm chromatin.

The DNA in human sperm chromatin is packaged into nucleoprotamine (approximately 85%) and nucleohistone (approximately 15%). Whether these two chromatin fractions are sequence-specific subsets of the spermatozoon genome is the question addressed in this report. Sequence-specific packaging would suggest distinct structural and functional roles for the nucleohistone and nucleoprotamine in late spermatogenesis or early development or both. After removal of histones with 0.65M NaCl, exposed DNA was cleaved with Bam HI restriction endonuclease and separated by centrifugation from insoluble nucleoprotamine. The DNA sequence distribution of nucleohistone DNA in the supernatant and nucleoprotamine DNA in the pellet was compared by cloning size-selected single-copy sequences and by using the derived clones as probes of nucleohistone DNA and nucleoprotamine DNA. Two clones derived from nucleohistone DNA preferentially hybridized to nucleohistone DNA, and two clones derived from nucleoprotamine DNA preferentially hybridized to nucleoprotamine DNA, which demonstrated the existence of sequence-specific nucleohistone and nucleoprotamine components within the human spermatozoon.

Premature p34cdc2 activation required for apoptosis.

Activation of the serine-threonine kinase p34cdc2 at an inappropriate time during the cell cycle leads to cell death that resembles apoptosis. Premature activation of p34cdc2 was shown to be required for apoptosis induced by a lymphocyte granule protease. The kinase was rapidly activated and tyrosine dephosphorylated at the initiation of apoptosis. DNA fragmentation and nuclear collapse could be prevented by blocking p34cdc2 activity with excess peptide substrate, or by inactivating p34cdc2 in a temperature-sensitive mutant. Premature p34cdc2 activation may be a general mechanism by which cells induced to undergo apoptosis initiate the disruption of the nucleus.

Rapid characterization of DNA oligomers and genotyping of single nucleotide polymorphism using nucleotide-specific mass tags.

Using currently available MS-based methods, accurate mass measurements are essential for the characterization of DNA oligomers. However, there is a lack of specificity in mass peaks when the characterization of individual DNA species in a mass spectrum is dependent solely upon the mass-to-charge ratio (m/z). Here, we utilize nucleotide-specific tagging with stable isotopes to provide internal signatures that quantitatively display the nucleotide content of oligomer peaks in MS spectra. The characteristic mass-split patterns induced by the partially (13)C/(15)N-enriched dNTPs in DNA oligomers indicate the number of labeled precursors and in turn the base substitution in each mass peak, and provide for efficient SNP detection. Signals in mass spectra not only reflect the masses of particular DNA oligomers, but also their specific composition of particular nucleotides. The measurements of mass tags are relative in the mass-split pattern and, hence, the accuracy of the determination of nucleotide substitution is indirectly increased. For high sample throughput, (13)C/(15)N-labeled sequences of interest have been generated, excised in solution and purified for MS analysis in a single-tube format. This method can substantially improve the specificity, accuracy and efficiency of mass spectrometry in the characterization of DNA oligomers and genetic variations.

Control of radiation-induced G1 arrest by cell-substratum interactions.

Ionizing radiation has been reported to cause an irreversible, senescence-like G1 arrest in human fibroblasts, which is accompanied by elevated p21CIP1 amounts. In further support of a senescence-like arrest, we show that expression of p53 and cyclin D1 is elevated in gamma-irradiated, arrested fibroblasts. However, we also demonstrate that the arrest is reversible if the irradiated cells are trypsinized and replated, which may implicate cellular-extracellular matrix interactions in cell cycle control after irradiation.

Differential regulation of cyclin-dependent kinase inhibitors in monolayer and spheroid cultures of tumorigenic and nontumorigenic fibroblasts.

The Ras proto-oncogene has been implicated in the in vivo development of tumors and in the in vitro transformation of cultured cell lines. In both of these conditions, Ras-mediated disruption of cell cycle-regulatory mechanisms leads to unregulated cellular proliferation, although the exact mechanisms by which Ras accomplishes this are not clear. Using as a model the M1 and MR1 rat fibroblast cell lines, which differ in the expression of a regulated Ras (M1 cells) versus a constitutively active Ras (MR1 cells), we examined the role of Ras in the control of cellular proliferation in two-dimensional (monolayer) and three-dimensional (spheroid) cell cultures. These cell lines are very similar in their monolayer growth characteristics, but M1 cells will arrest their cell cycle progression in aggregate culture, whereas MR1 cells proliferate normally as small spheroids. We report here that G1-phase arrest in plateau-phase monolayer cultures of both M1 and MR1 cells correlates with up-regulated expression of the cyclin-dependent kinase (CDK) inhibitor p18INK4c. Enhanced p18INK4c expression was also observed in G1-arrested M1 cells cultured as multicellular spheroids but was not induced in small proliferating MR1 multicellular spheroids. The kinetics of G1 arrest in M1 cells after inoculation into aggregate culture correlated well with the induction of p18INK4c expression. Conversely, resumption of proliferation in monolayer culture of arrested M1 cells isolated from spheroids coincided with the loss of expression of p18INK4c. After extended culture, cells in the inner region of MR1 spheroids arrested in the G1 phase without any up-regulation of p18INK4c expression. In this case, the CDK inhibitor p21(Cip1/Waf1) was selectively induced in the inner regions of large MR1 spheroids, concomitant with a decrease in cyclin and CDK expression. Thus, Ras-dependent regulation of p18INK4c expression seems to control the ability of rat embryo fibroblasts to proliferate as small multicellular aggregates, whereas p21(Cip1/Waf1) expression seems to regulate the G1-phase arrest induced by the stressful microenvironment found within the inner region of large spheroids.

The kinase inhibitor staurosporine induces G1 arrest at two points: effect on retinoblastoma protein phosphorylation and cyclin-dependent kinase 2 in normal and transformed cells.

Staurosporine (ST), a protein kinase inhibitor, at a concentration of 20 nM arrests normal diploid fibroblasts 3 h into G1 (H. A. Crissman et al., Proc. Natl. Acad. Sci. USA, 88: 7580-7584, 1991; K. Abe et al., Exp. Cell Res., 192: 122-127, 1991). ST (2 nM) induces a new G1 arrest point at 6 h into G1. Partial phosphorylation of the retinoblastoma protein was observed at the 2 nM ST arrest point, whereas the retinoblastoma protein was unphosphorylated or underphosphorylated at the 20 nM arrest point. This correlated with the activity of the cyclin-dependent kinase 2 (CDK2) and the phosphorylation of the Thr160 residue of p33CDK2. The cyclin E and cyclin D1/2 levels were reduced at the 20 nM ST arrest point. In HeLa cells that do not arrest in G1 in response to 2 or 20 nM ST, the retinoblastoma protein and CDK2 phosphorylations and CDK2 activity were not affected by ST. These results suggest that ST inhibits one or more G1-regulating protein kinases, which lie upstream of CDK2.

Immortalization of human fibroblasts by SV40 large T antigen results in the reduction of cyclin D1 expression and subunit association with proliferating cell nuclear antigen and Waf1.

Protein complexes containing cyclins and cyclin-dependent protein kinases (cdks) have been shown to be rearranged in both spontaneous and viral tumor antigen-transformed cells. We have examined G1- and S-phase cyclin/cdk complexes as a function of the neoplastic progression of human diploid fibroblasts transfected with the SV40 large T antigen. We find that the expression of cyclin D1 and its association with proliferating cell nuclear antigen (PCNA) and Waf1 remain unchanged in precrisis human fibroblasts transfected with SV40 large T antigen. However, in these same cells the association of cdk4 with cyclin D1, PCNA, and Waf1 is disrupted. Upon immortalization, cyclin D1 protein expression is decreased, and binding of both PCNA and Waf1 with the remaining cyclin D1 is reduced. In contrast, large T antigen increased the expression of cyclin A and cyclin E proteins in both precrisis and immortal cells and did not reduce the binding of PCNA or Waf1 to either cdk2 or cyclin A proteins. These results show that large T-antigen expression in human fibroblasts selectively uncouples cyclin D1 from cdk4, and subsequent immortalization of these cells results in additional changes to the cyclin D1-dependent cell cycle regulatory pathways.

A brief staurosporine treatment of mitotic cells triggers premature exit from mitosis and polyploid cell formation.

At any point during the progression of many tumor types, cells can develop a hyperploid DNA content. Hyperploid tumors are significant more aggressive, with a higher growth rate and a poor patient prognosis. Yeast genetics have implicated three important genes involved in DNA ploidy changes: cdc2, cyclin b, and a specific inhibitor of the p34(cdc2)/cyclin B kinase, rum1. Mutations in these genes uncoupled the dependence mitosis on DNA replication in the fission yeast, Saccharomyces pombe. It was proposed that the inactivation of the mitotic kinase complex, p34(cdc2)/cyclin B, induces a G(1), state wherein the cells re-replicate their DNA without an intervening mitosis. We show in this report that treatment of only M phase-arrested mouse cells, with the protein kinase inhibitor staurosporine, induced polyploidy. Nocodazole-arrested metaphase FT210 cells were pulsed with 100 ng/ml of staurosporine for 1 h. This 1-h treatment results in the inhibition of the mitotic p34(cdc2) kinase. The inhibition of the mitotic kinases leads to a reduction in the histone H1 and H3 mitotic-associated phosphorylations, chromosome decondensation and nuclear membrane reformation. When released into normal growth medium, these cells are reset to a G(1)state, re-replicate their DNA without completing mitosis, and become octaploid.

Requirements for p53 and the ATM gene product in the regulation of G1/S and S phase checkpoints.

We investigated the requirements for protein p53 and the ATM gene product in radiation-induced inhibition of DNA synthesis and regulation of the cyclin E/ and cyclin A/cyclin dependent kinases (Cdks). Wild type (WT) mouse lung fibroblasts (MLFs), p53(-/-) knock-out MLFs, normal human skin fibroblasts (HSF-55), and human AT skin fibroblasts (GM02052) were used in the investigations. The absence of p53 had no significant effect on the inhibition or recovery of DNA synthesis throughout the S phase, as determined from BrdU labeling and flow cytometry, or the rapid inhibition of cyclin A/Cdks. Gamma radiation (8 Gy) inhibited DNA synthesis and progression into G2 during the first 3 h after irradiation, and the recovery of these processes occurred at similar rates in both WT and p53(-/-) MLFs. The cyclin A/Cdks were inhibited 55-70% at 1 h after irradiation in both cell types, but p21WAF1/Cip1 levels or p21 interaction with Cdk2 did not increase in the irradiated p53(-/-) MLFs. Although p53(-/-) MLFs do not exhibit prolonged arrest at a G1 checkpoint, radiation did induce a rapid 20% reduction and small super-recovery of cyclin E/Cdk2 within 1-2 h after irradiation. Similar inhibition and recovery of cyclin E/Cdk2 previously had been associated with regulation of transient G1 delay and the inhibition of initiation at an apparent G1/S checkpoint in Chinese hamster cells. In contrast, loss of the ATM gene product abrogated transient cyclin E/Cdk2 inhibition, most inhibition of DNA synthesis and all, but a 10-15% inhibition, of the cyclin A/Cdks. The results indicate that neither p53 nor p21 is required for transient inhibition of cyclin E/Cdk2 associated with the G1/S checkpoint or for inhibition of DNA synthesis at 'checkpoints' within the S phase. Conversely, the ATM gene product appears to be essential for regulation of the G1/S checkpoint and for inhibition of DNA replication associated with the inhibition of cyclin A/Cdk2. Differential aspects of DNA synthesis inhibition among cell types are presented and discussed in the context of S phase checkpoints.

The conformation of histone H5 bound to DNA. Maintenance of the globular structure after binding.

Trypsin digestion is used to investigate the conformation of histone H5 when bound to DNA. A central region of H5 comprising residues (22--100) is found to be resistant to digestion and it is concluded that this region is compacted whilst the remaining N- and C-terminal regions are more extended. Since this is the same result found previously for the free solution conformation of histone H5 it follows that a 3-domain structure is preserved on DNA binding. The binding of H5 and the central region (22--100) to DNA is also studied using proton magnetic resonance (270 MHz) and a precipitation approach. It is concluded that all 3 domains of H5 bind to DNA at low ionic strengths. The central domain (residues 22--100) is released at 0.3--0.4 M NaCl, but 0.7 M NaCl is required to release the N- and C-terminal regions. Comparison is made of H5 binding to DNA with that of the related histone H1.

Chromosomal loop/nuclear matrix organization of transcriptionally active and inactive RNA polymerases in HeLa nuclei.

The relative distribution of transcriptionally active and inactive RNA polymerases I and II between the nuclear matrix/scaffold and chromosomal loops of HeLa cells was determined. Total RNA polymerase was assessed by immunoblotting and transcribing RNA polymerase by a photoaffinity labeling technique in isolated nuclei. Nuclear matrix/scaffold was isolated by three methods using high-salt, intermediate-salt or low-salt extraction. The distribution of RNA polymerases I and II were very similar within each of the methods, but considerable differences in distributions were found between the different preparation methods. Either intermediate-salt or high-salt treatment of DNase I-digested nuclei showed significant association of RNA polymerases with the nuclear matrix. However, intermediate-salt followed by high-salt treatment released all transcribing and non-transcribing RNA polymerases. Nuclear scaffolds isolated with lithium diiodosalicylate (low-salt) contained very little of the RNA polymerases. This treatment, however, caused the dissociation of RNA polymerase II transcription complexes. These results show unambiguously that RNA polymerases, both in their active and inactive forms, are not nuclear matrix proteins. The data support models in which the transcriptional machinery moves around DNA loops during transcription.

Mobility of positioned nucleosomes on 5 S rDNA.

We report on a dynamic aspect of nucleosome positioning, in the absence of transcription-related events, on sea urchin 5 S rDNA. On tandem repeats of nucleosome length DNA of this strongly positioning sequence, histone octamers assemble in one dominant position surrounded by minor positions, ten base-pairs apart and therefore with identical rotational setting of the DNA coil. The existence of this cluster of positions, determined using micrococcal nuclease is confirmed by the results from DNase I footprinting and restriction enzyme analysis. The results from these techniques and from two-dimensional nucleoprotein polyacrylamide gel analysis indicate that the cluster of octamer positions is in dynamic equilibrium, in low ionic conditions, suggesting that the minor positions reflect fluctuations around the major nucleosome site. Histone octamer mobility appears to be temperature dependent and is reversibly inhibited by Mg2+.

Chromatosome positioning on assembled long chromatin. Linker histones affect nucleosome placement on 5 S rDNA.

Long chromatin containing linker histones H1 or H5 was assembled on tandemly repeated 172 or 207 base-pair nucleosome positioning sequences from a sea urchin 5 S RNA gene. The effects of H1 and H5 on spacing and positioning of nucleosomes were assessed. In the absence of linker histones, precise determinations of core particle boundaries showed that, although a large proportion of the histone octamers occupy a unique position, there is a small group of other, less populated sites located around this major site. The dominant position was found 10 to 15 base-pairs upstream from the unique position previously reported for the histone octamer on the monomer 260 base-pair sequence. Linker histones do not override the underlying DNA signals that induce the very regular spacing of nucleosomes in chromatins assembled on these strongly positioning multimer DNA sequences. They were nevertheless found to be decisive in determining the chromatosome positions and their distributions, and as such define the chromatosome as a positioning entity.

Methylation is co-ordinated on the putative replication origins of Physarum ribosomal DNA.

In Physarum polycephalum, the ribosomal DNA is found as 60,000 base-pair palindromes. Each rDNA has four symmetrically arranged replication origins flanked by ribosomal RNA genes. A particular sequence, the putative replication origin, is repeated at the approximate position of each origin and nowhere else in the molecule. On a typical rDNA molecule, only one origin is active per replication cycle. We show that both the level and co-ordination of methylation result in asymmetrically methylated rDNA molecules that are particularly hypomethylated at one of their four putative replication origins. This pattern of methylation on a typical rDNA molecule is consistent with a model where hypomethylation is a determinant of origin activity.

Nucleosome arrays inhibit both initiation and elongation of transcripts by bacteriophage T7 RNA polymerase.

We have examined the effects of nucleosome cores on the initiation and elongation of RNA transcripts by phage T7 RNA polymerase in vitro. A transcription template, pT207-18, was constructed containing tandemly repeated 207 base-pair (bp) nucleosome positioning sequences from a sea urchin (Lytechinus variegatus) 5 S RNA gene inserted between the T7 and SP6 transcription promoters of pGEM-3Z. Nucleosome cores were reconstituted onto supercoiled, closed circular pT207-18 DNA and double label transcription experiments were performed to determine the effects of nucleosome cores on the initiation and elongation of transcripts by T7 RNA polymerase. Both transcript initiation and elongation were inhibited, the extent of the inhibition being directly proportional to the number of nucleosome cores reconstituted onto the pT207-18 DNA templates. Time course transcription experiments indicated that nucleosome cores caused a reduction in the equilibrium length of transcripts and not mere retardation of elongation rates. Continuous regularly spaced linear arrays of nucleosomes were obtained by digesting reconstituted nucleosomel pT207-18 templates with DraI, for which a unique restriction site lies within the nucleosome positioning region of the 207 bp 5 S rDNA repeat sequence. After in vitro transcription with T7 RNA polymerase an RNA ladder with 207 nucleotide spacing was obtained, indicating that transcription can occur through continuous arrays of positioned nucleosome cores. It is demonstrated that nucleosome cores partially inhibit the elongation of transcripts by T7 RNA polymerase, while allowing passage of the transcribing polymerase through each nucleosome core at an upper limit efficiency of 85%. Hence, complete transcripts are produced with high efficiency from short nucleosomal templates, while the production of full-length transcripts from long nucleosomal arrays is relatively inefficient. The results indicate that nucleosome cores have significant inhibitory effects in vitro not only on transcription initiation but on transcription elongation as well, and that special mechanisms may exist to overcome these inhibitory effects in vivo.

Fragile X DNA triplet repeats, (GCC)n, form hairpins with single hydrogen-bonded cytosine.cytosine mispairs at the CpG sites: isotope-edited nuclear magnetic resonance spectroscopy on (GCC)n with selective 15N4-labeled cytosine bases.

Here, we provide a direct proof that the formation of hairpins by (GCC)n at the 5'-UTR of the FMR-1 gene offers a mechanism for CpG hypermethylation associated with the fragile X syndrome. For this, we have performed hetero-nuclear (15N-1H) magnetic resonance spectroscopy to probe the structure of the CpG sites in the (GCC)n hairpins that are 15N-labeled at the amino (N4) groups of specific cytosine bases. Analyses of chemical shift, pH-induced chemical exchange, and NOE pattern of the (15N-labeled) amino protons of cytosine bases reveal that the cytosine bases at the CpG sites are intrahelical and well-stacked with the neighboring G.C base-pairs in the stem of these hairpins and probably form single hydrogen-bonded C.C mispairs. Measurements of pH-dependent 1H line-width also demonstrate that the C.C mispairs are more susceptible to open-closure than the G.C base-pairs. Thus, the Cs at the CpG sites of the (GCC)n hairpin are "flipped out" more easily to the activated state than those in the corresponding Watson-Crick duplex, (GCC)n. (GGC)n and this makes the hairpin a better target for methylation by the human methyltransferase, the enzyme that methylates the Cs at the CpG sites.

Structure-function correlations of the insulin-linked polymorphic region.

The insulin minisatellite of the insulin-linked polymorphic region (ILPR), a 14 base-pairs long tandem repeat of: 5'-ACAGGGGTGTGGGG-3' 3'-TGTCCCCACACCCC-5', is located 363 base-pairs upstream of the human insulin gene. A locus for insulin-dependent diabetes mellitus (IDDM) has been mapped to the ILPR. It has been shown that the ILPR is polymorphic in length and this length polymorphism is also related to the transcriptional activity of the insulin gene and the susceptibility to IDDM. Here, we attempt to decipher the role of the ILPR structure in length polymorphism and transcriptional regulation. We show by gel electrophoresis, circular dichroism (CD) and one and two-dimensional nuclear magnetic resonance spectroscopy (1D/2D NMR) that the G-rich strand of the ILPR adopts an intramolecularly folded hairpin G-quartet structure. A detailed analysis of 1D/2D NMR data of d(G4TGTG4) and d(G4TGTG4ACAG4TGTG4) enables us to define the nature of chainfolding, the stacking interaction of the G-tetrads in the stem, and the interactions of the bases in the loops. d(G4TGTG4ACAG4TGTG4) happens to be the smallest unit of the G-rich strand that can form the intramolecular hairpin G-quartet structure. For long ILPR sequences, several such hairpin G-quartet structures can be linked in space. Indeed, by an in vitro replication assay, we show the presence of such multiple hairpin G-quartet structures for the G-rich strand of the ILPR of repeat length 6. This observation suggests that the formation of multiple hairpin G-quartets may explain slippage during replication and the observed length polymorphism. From our high resolution structure, we are able to identify a set of interactions that are critical for the structure and stability of the hairpin G-quartet. Single or double mutations in the ILPR that destabilize these interactions also lower the transcriptional activity of the insulin gene. Therefore, the hairpin G-quartet structure of the ILPR has a direct correlation with the transcriptional activity of the human insulin gene.