Antisense oligonucleotides - the way forward

Antisense oligonucleotides provide attractive possibilities for developing a new class of drugs and the design principle involves straightforward base-pairing rules. The approach has already been useful in many functional genomics efforts and there has also been recent clinical success. With this progress in place, it now is appropriate to thoroughly address design uncertainties and unforeseen effects that have emerged. The key objective has long been, and remains to be, the identification of novel oligonucleotide analogs providing the possibility of achieving high in vivo efficacy in the absence of significant toxicity; such compounds are not at hand today but certain advances have been made in recent years. There is also a need for more expertise in target site selection and for improving delivery and/or bioavailability properties. Immediately achievable goals in the field implicate use of improved in vitro and in vivo assays and, importantly, a set of standard controls, uniform between laboratories, such that results can be compared in a rational manner. In turn, this should lead to the emergence of antisense agents that are useful for a wide range of research and therapeutic applications.

Essential structural features in the Schizosaccharomyces pombe pre-rRNA 5' external transcribed spacer

The proximal region in the 5' external transcribed spacer (5'ETS) of the genes encoding ribosomal RNAs in Schizosaccharomyces pombe was examined with respect to structural features which underlie rRNA maturation. Computer analyses and partial digestion with nuclease probes indicate a crucifix-like structure composed primarily of three extended hairpins which are more highly ordered than previously proposed in Saccharomyces cerevisiae. A re-evaluation of the same region in S. cerevisiae indicates a conserved core structure, including the U3 snoRNA binding site within this higher-order structure. The sequences encoding the individual hairpins were deleted by PCR-mediated mutagenesis and the mutant rDNAs were expressed in vivo to determine the effect of these features on rRNA maturation. Quantitative hybridization analyses indicate that the first hairpin only has modest effects on 18 S rRNA maturation, but the other two regions are critical and no mature 18 S rRNA was observed. When smaller changes were systematically introduced into the critical regions, strong correlations were observed with known or putative events in rRNA maturation. Changes associated with an intermediate cleavage site in helix II and with the putative U3 snoRNA binding site were again critical to 18 S rRNA production. In each case, the effects were sequence dependent and not simply the result of disrupted structure. Further analyses of the 5.8 S rRNA indicate that the large ribosomal subunit RNA can be properly processed in each case but the efficiency is reduced by as much as 60 %, an observation which provides new evidence of interdependency in the maturation process. The results illustrate that rRNA processing is more critically dependent on the 5'ETS than previously believed.

Gene expression profiles in HTLV-I-immortalized T cells: deregulated expression of genes involved in apoptosis regulation

Human T-cell leukemia virus type I (HTLV-I) is the etiologic agent of adult T-cell leukemia, an acute and often fatal T-cell malignancy. A key step in HTLV-I-induced leukemigenesis is induction of abnormal T-cell growth and survival. Unlike antigen-stimulated T cells, which cease proliferation after a finite number of cell division, HTLV-I-infected T cells proliferate indefinitely (immortalized), thus facilitating occurrence of secondary genetic changes leading to malignant transformation. To explore the molecular basis of HTLV-I-induced abnormal T-cell survival, we compared the gene expression profiles of normal and HTLV-I-immortalized T cells using 'gene array'. These studies revealed a strikingly altered expression pattern of a large number of genes along with HTLV-I-mediated T-cell immortalization. Interestingly, many of these deregulated genes are involved in the control of programmed cell death or apoptosis. These findings indicate that disruption of the cellular apoptosis-regulatory network may play a role in the HTLV-I-mediated oncogenesis.

Postural control--a comparison between patients with chronic anterior cruciate ligament insufficiency and healthy individuals

Postural control in the sagittal plane was evaluated in 22 patients with chronic anterior cruciate ligament (ACL) deficiency and the result was compared to that of a control group of 20 uninjured subjects. Measurement of the body sway was done on a fixed and sway-referenced force plate in both single-limb and two-limb stance, with the eyes open and closed, respectively. Further, an analysis of the postural reactions to perturbations backwards and forwards, respectively, was made in single-limb stance. The results demonstrated statistically significant deficits of the postural control in the patient group compared to the control group, but also within the patient group. There was a significantly higher body sway within the patient group when standing on a stable support surface on the injured limb than standing on the uninjured limb with the eyes open, but no difference with the eyes closed. When standing on a stable support surface, there was a significantly higher body sway in the patient group standing on the injured leg than in the control group, both with eyes open and closed. The patient group also showed a significantly impaired postural control compared to the control group when standing on the uninjured leg with the eyes closed. There was no difference between the groups in the two-limb stance. When standing on the sway-referenced support surface, the patient group had a significantly larger body sway than the control group when the eyes were open, but there was no significant difference between the groups with the eyes closed. The measurement of the postural corrective responses to perturbations backwards and forwards showed that the reaction time measured from the initiation of the force plate translation, and the amplitude of the body sway was significantly greater in the patient group than in the control group. We conclude that patients with a continuing chronic ACL insufficiency several years after injury have an impaired postural control in the antero-posterior direction in single-limb stance on their injured leg. They also show a greater body sway and a prolonged reaction time when subjected to antero-posterior perturbations when standing on their injured leg.

NF-kappaB-inducing kinase and IkappaB kinase participate in human T-cell leukemia virus I Tax-mediated NF-kappaB activation

The tax gene product of human T-cell leukemia virus I induces aberrant expression of various cellular genes, which contributes to transformation of host cells. Induction of many Tax target genes is mediated through transcription factor NF-kappaB. Here we show that Tax triggers activation of cellular protein kinases, IkappaB kinase alpha (IKKalpha) and IKKbeta, which phosphorylate the NF-kappaB inhibitory protein IkappaB alpha, resulting in its degradation and NF-kappaB activation. Constitutive IKK activation occurs in both Tax-transfected and human T-cell leukemia virus I-infected T cells. We further demonstrate that Tax-mediated NF-kappaB signaling also requires the NF-kappaB-inducing kinase (NIK). Consistently, inactive forms of either IKKs or NIK attenuate Tax-mediated NF-kappaB activation. Therefore, Tax activates NF-kappaB by targeting cellular signaling molecules, including both IKKs and NIK.

Antisense inhibition of gene expression in bacteria by PNA targeted to mRNA

Peptide nucleic acid (PNA) is a DNA mimic with attractive properties for developing improved gene-targeted antisense agents. To test this potential of PNA in bacteria, PNAs were designed to target the start codon regions of the Escherichia coli beta-galactosidase and beta-lactamase genes. Dose-dependent and specific gene inhibition was observed in vitro using low nanomolar PNA concentrations and in vivo using low micromolar concentrations. Inhibition was more efficient for a permeable E. coli strain relative to wild-type K-12. The potency of the anti-beta-lactamase PNAs was abolished by a six base substitution, and inhibition could be re-established using a PNA with compensating base changes. Antisense inhibition of the beta-lactamase gene was sufficient to sensitize resistant cells to the antibiotic ampicillin. The results demonstrate gene- and sequence-specific antisense inhibition in E. coli and open possibilities for antisense antibacterial drugs and gene function analyses in bacteria.

Inhibition of translation and bacterial growth by peptide nucleic acid targeted to ribosomal RNA

Peptide nucleic acid (PNA) is a DNA mimic that has shown considerable promise as a lead compound for developing gene therapeutic drugs. We report that PNAs targeted to functional and accessible sites in ribosomal RNA can inhibit translation in an Escherichia coli cell-free transcription/translation system, with 50% reductions caused by nanomolar PNA concentrations. The effect in vitro is quantitatively similar to that of the known translation inhibitor and antibiotic tetracycline. Also, the targeted PNAs inhibited bacterial growth on agar plates and in liquid culture. A strain of E. coli (AS19) that is more permeable to antibiotics was approximately 10-fold more sensitive to the active PNAs, suggesting that the effect on growth indeed was caused by PNAs that entered cells. Inhibition was not observed when using control PNAs of similar composition but with an unrelated or mismatched sequence. The results demonstrate that ribosomal RNA is a possible target for sequence-designed novel antibiotics based on DNA analogues or mimics.

Interdependence in the processing of ribosomal RNAs in Schizosaccharomyces pombe

Eukaryotic rRNAs are produced by cleavage of a large 35 to 45 S pre-rRNA transcript which initially must be fully transcribed and assembled into an 80 to 90 S nucleolar ribonucleoprotein particle. Despite this need for a completed transcript, several investigations have reported a split processing scheme for independent maturation of the large and small subunit rRNAs. Here, an efficiently expressed rDNA plasmid was used to quantitatively analyze the effects of mutations in the internal transcribed spacer (ITS) region in the yeast, Schizosaccharomyces pombe. The results show that substitution of ITS regions inhibits the processing of distant external transcribed spacers (ETS) and that deletion of the ITS2 spacer not only prevents the maturation of the large subunit, but severely affects maturation of the small subunit rRNA. This indicates that the processing mechanisms are not fully split and, when taken together with other evidence of interdependences in rRNA maturation, the results suggest that the interdependences act as a quality control mechanism to help ensure that only functional rRNA is incorporated into ribosomes.

Progress in developing PNA as a gene-targeted drug

Peptide nucleic acid (PNA) is a DNA mimic in which the nucleobases are attached to a pseudopeptide backbone. This achiral, uncharged, and rather flexible peptide backbone permits more stable hybridization to DNA and RNA oligomers with uncompromised or even improved sequence selectivity. Additional advantages of PNA are stability against nucleases and proteases and convenient solid phase synthesis. At the RNA level, PNA can be targeted to mRNA to block protein synthesis in an antisense strategy. PNA can also be targeted to the RNA component of ribonucleoproteins (RNPs) to inhibit their enzymatic activities. At the DNA level, the unique ability of PNA to bind DNA by duplex invasion can be used to arrest transcription within a gene sequence or to provide an artificial open complex to promote transcription. This review focuses on recent progress toward the development of PNA as a sequence-targeted drug.

The ribosomal-RNA-processing pathway in Schizosaccharomyces pombe

In all cells, a long precursor RNA is processed into mature rRNAs for ribosome biogenesis. In eukaryotes, the complexity and speed of the overall process often has made it difficult to establish finer details of the maturation pathway. Since phylogenetic comparisons can provide evidence for critical events, the major rRNA processing pathway for the yeast Schizosaccharomyces pombe was determined using primer extension, nuclease protection and Northern-hybridisational analyses. Transcript mapping of the 5' external transcribed spacer revealed six cleavage sites which occur upstream of the mature 18S termini. Two of these sites as well as a site adjacent to the 18S termini are complementary to conserved Box sequences in the S. pombe U3 small nucleolar RNA. Transcript mapping of the internal transcribed spacers (1 and 2) suggest similar maturation schemes for the two spacers, in which an initial endonuclease cleavage is followed by processing to the mature termini. The mature 5' termini of 25S rRNA appear to be heterogeneous in S. pombe, as has been demonstrated for 5.8S rRNA, suggesting an essential limiting structure in the ribosome-integrated mature RNA. Together with our previous analysis of the 3' external spacer region, the results reveal the major processing pathway for S. pombe and further support a maturation process which acts as a quality assurance mechanism.

Constitutive dephosphorylation and activation of a member of the nuclear factor of activated T cells, NF-AT1, in Tax-expressing and type I human T-cell leukemia virus-infected human T cells

The tax gene product of the type I human T-cell leukemia virus (HTLV-I) transactivates interleukin-2 (IL-2) gene through activation of an enhancer termed CD28 responsive element (CD28RE). Tax activation of the CD28RE is partially mediated by a member of the nuclear factor of activated T cells, NF-AT1. We have previously shown that NF-AT1 is constitutively active in Jurkat T cells stably transfected with the Tax cDNA, although the underlying molecular mechanism and physiological relevance of this finding remain unclear. In this report, we demonstrate that the active form of NF-AT1 is also present in the nuclei of HTLV-I-transformed T cells that express the Tax protein. Interestingly, the constitutive activation of NF-AT1 in these T cells is associated with its dephosphorylation. Furthermore, the dephosphorylated NF-AT1 can be rapidly rephosphorylated when the cells are incubated with cyclosporin A, an immunosuppressant inhibiting the serine/threonine phosphatase calcineurin. These results suggest that activation of NF-AT1 in Tax-expressing and HTLV-I-transformed T cells results from its dephosphorylation, which in turn may be due to deregulation of calcineurin.

CD28 mediates a potent costimulatory signal for rapid degradation of IkappaBbeta which is associated with accelerated activation of various NF-kappaB/Rel heterodimers

Optimal activation of T cells requires at least two signals delivered by the T-cell receptor complex and costimulatory molecules such as CD28. The CD28 signaling participates in the transcription of the interleukin-2 gene through activation of an enhancer termed the CD28-responsive element (CD28RE). Stimulation of CD28 enhances mitogen-mediated induction of CD28RE-binding proteins including members of the NF-kappaB/Rel transcription factor family, although the underlying mechanism remains elusive. In this report, we show that CD28 costimulation leads to biphasic induction of NF-kappaB/Rel heterodimers, including early-phase induction of p50/RelA and c-Rel/RelA and late-phase induction of p50/c-Rel. Interestingly, activation of these NF-kappaB/Rel complexes by the CD28 signal is associated with the rapid degradation of both IkappaBalpha and IkappaBbeta, two major cytoplasmic inhibitors of NF-kappaB/Rel. Although IkappaBalpha degradation can be induced by phorbol ester alone, degradation of IkappaBbeta is largely dependent on the CD28 costimulatory signal. We further demonstrate that CD28-mediated transactivation of the CD28RE enhancer is potently inhibited by an N-terminal truncation mutant of IkappaBbeta that is incapable of responding to the degradation signals. Together, these results suggest that the CD28 costimulatory signal augments activation of NF-kappaB/Rel by promoting degradation of IkappaBbeta as well as enhancing degradation of IkappaBalpha and that induction of NF-kappaB/Rel serves as an essential step in the signal-mediated activation of the CD28RE enhancer.

Regulation of dihydrofolate reductase gene expression and E2F components in human diploid fibroblasts during growth and senescence

The induction of dihydrofolate reductase (DHFR), a key enzyme in DNA biosynthesis that is induced just before the onset of S phase, is markedly attenuated in senescent human fibroblasts (Pang and Chen, 1994, J. Cell. Physiol., 160:531-538). Footprinting analysis of the 365 bp promoter region of the human DHFR gene (-381 to -17) indicated that nuclear proteins bind to a cluster of cis-elements, including two overlapping E2F binding sequences, two Sp1 sites, and one Yi sequence. Gel mobility shift assays were performed to assess the role of each cis-element in the regulation of DHFR gene expression. We found that 1) Sp1 binding activity was constitutively expressed throughout the cell cycle in early passage and senescent cells; 2) Yi binding activity was undetectable in both early passage and senescent cells; and 3) E2F binding activity was serum-inducible, senescence-dependent, and prominent in presenescent cells but strikingly diminished in senescent cells. Northern blot analysis of the expression of E2F and DP family members showed that the E2F-1, E2F-4, and E2F-5 mRNA was growth- and senescence-dependent, whereas E2F-3, DP-1, and DP-2 expression was constitutive and senescence-independent. In contrast, E2F-2 mRNA was not detectable in IMR-90 or WI-38 human fibroblasts. Western blot analysis showed that among the E2F-associated proteins, the expression of E2F-1, cyclin A, and cyclin B but not p107 was cell cycle- and senescence-dependent. A nuclear extract mixing experiment suggested that an inhibitory factor may further reduce E2F binding activity in senescent cells.

Activation of the IL-2 gene promoter by HTLV-I tax involves induction of NF-AT complexes bound to the CD28-responsive element

The tax gene product of the type I human T-cell leukemia virus (HTLV-I) is a potent transcriptional activator of various growth-related cellular genes, including that encoding interleukin-2 (IL-2). Tax activation of many of these target genes appears to be mediated by the NF-kappa B/Rel and CREB/ATF family of cellular transcription factors. However, the mechanism by which Tax transactivates the IL-2 gene remains unclear. In the present study, we demonstrate that neither NF-kappa B/Rel nor CREB/ATF is sufficient for Tax-mediated activation of the IL-2 promoter. Two novel nuclear protein complexes are induced by Tax and specifically bind to an IL-2 gene enhancer, the CD28-responsive element (CD28RE). Immunobiochemical analyses suggest that these DNA binding complexes contain at least two members of the nuclear factor of activated T cells, NF-ATp and NF-ATc. However, the CD28 binding NF-AT complexes do not contain Jun and Fos family proteins that have been proposed to serve as NF-AT partners in the activation of the IL-2 NF-AT motif. Transient transfection studies demonstrate that the in vivo expressed NF-ATp binds to the CD28RE probe and enhances Tax-mediated activation of this critical IL-2 enhancer. We demonstrate further that binding of NF-AT to CD28RE is critical for Tax activation of the IL-2 promoter. Together, these results suggest a novel mechanism of Tax-mediated activation of the IL-2 gene, which involves the induction of NF-AT-containing CD28RE binding complexes.

Persistent activation of NF-kappa B/Rel by human T-cell leukemia virus type 1 tax involves degradation of I kappa B beta

Activation of the eukaryotic NF-kappaB/Rel transcription factors by various cytokines and mitogens is a transient event, reflecting the fact that these inducers trigger the degradation and resynthesis of the dynamic NF-kappaB/Rel inhibitor IkappaBalpha. However, the tax gene product of the human T-cell leukemia virus type 1 (HTLV-1) is known to induce the persistent nuclear expression of various NF-kappaB/Rel factors, especially the c-Rel proto-oncoprotein, although the underlying mechanism remains unclear. In the present study, we demonstrate that Tax induces the degradation Of IkappaBbeta, another NF-kappaB/Rel cytoplasmic inhibitor that differs from IkappaBalpha in signal responses. Unlike that observed with IkappaBalpha, the degradation Of IkappaBbeta is not associated with its rapid resynthesis, apparently because of the failure of Tax to stimulate IkappaBbeta gene transcription. Thus, expression of Tax in Jurkat T cells leads to the gradual depletion of IkappaBbeta, which is correlated with the induction of c-Rel-containing kappaB binding complexes. Remarkably, in the three HTLV-1-infected T-cell lines investigated, little or no detectable amount of IkappaBbeta was found. We further demonstrate that Tax is able to override the cytoplasmic retention of c-Rel by 1kappaBbeta in transiently transfected cells. Together, these studies suggest that Tax-mediated inactivation Of IkappaBbeta may play a role in the persistent nuclear expression of c-Rel induced by HTLV-I infection.

An efficiently expressed 5.8S rRNA 'tag' for in vivo studies of yeast rRNA biosynthesis and function

Inefficient expression or detrimental markers have limited mutational analyses of eukaryotic 5.8S rRNA and the associated rDNA transcribed spacers. We have found a neutral, 4-base insertion mutation that effectively tags the 5.8S rRNA for improved studies of rRNA expression, processing and function. Cells expressing the tagged rDNA plasmid contain 50-60% mutant 5.8S rRNA, but show a normal growth rate and polysomal profile and a constant distribution of tagged 5.8S rRNA. The high level of expression also demonstrates that plasmid-associated rDNA is preferentially transcribed over chromosomal copies.

In-vitro correlation between tension and length change in an anterior cruciate ligament substitute

The length change and tension patterns from multiple insertion locations of an anterior cruciate ligament substitute were studied in 10 cadaver knees. Length change was measured with a spring-loaded isometer of low stiffness, and tension was measured with a piezoelectric load cell. In both instances a thin Kevlar test ligament was positioned in five different femoral and two different tibial ligament insertion locations, that were all located within the normal attachments of the anterior cruciate ligament. Differences were found regarding length changes and tension patterns from a simulated active extension between the central, posterior, and anterior femoral locations. All locations showed larger length change and tension values in extension than in flexion. The anterior femoral ligament insertion location showed length change and tension patterns with increasing values in flexion compared to the other femoral locations. The anterior tibial ligament insertion location showed smaller excursions of both length and tension, than did the central one, but the patterns of the curves were similar. A statistically significant correlation was found between length change and tension patterns throughout a 130-0 degrees range of motion. A statistically significant correlation was also found between the maximum length and tension values. No fixed relationship was found between the magnitude of the length and tension values, when different intervals of the range of motion were studied. RELEVANCE: The intraoperative employment of length change measurements of a test ligament in anterior cruciate ligament reconstruction gives information on where high tension can be expected in the range of motion of the knee, and how this can differ depending on the angle of graft fixation. The information gained can also be used to improve drill channel location. However, no predictions on the magnitude of tension can be made, mainly due to large biological variability.

Analysis of sequence-specific binding activity of cis-elements in human thymidine kinase gene promoter during G1/S phase transition

Expression of thymidine kinase (TK) gene in normal human diploid, cells is both cell cycle and age dependent and appears to be transcriptionally regulated. Several studies have indicated that the G1/S control sequence may reside within the region of about 130 bp upstream of the transcription initiation site. We have previously shown that a trans-acting factor, CBP/tk (CCAAT binding protein for TK gene), binds to either one of the two inverted CCAAT boxes in a cell cycle- and age-dependent manner (Pang and Chen, 1993, J. Biol. Chem., 268:2909-2916). An upstream 25 bp fragment (-109/-84), containing both Yi-like and E2F-like binding sites, has recently been proposed to be essential for the G1/S regulation of human TK gene. To assess the contribution of various cis-elements in human TK promoter to the G1/S regulation, we have examined the binding activity of these cis-elements in the nuclear extracts derived from human IMR-90 cells at low passage number. Our results indicated that no binding activity could be detected using either the 25 bp fragment (-109/-94) or the authentic Yi sequence. However, Yi binding activity was observed in SV-40 transformed IMR-90 cells. In contrast, the 28 bp fragment (-91/-64) that contains the distal inverted CCAAT box exhibited a strong binding in serum-stimulated young IMR-90 cells. The binding of CBP/tk to the 28 bp fragment was abolished by a single base mutation in the CCAAT box. The CBP/tk binding of the 28 bp fragment could not be displaced by either the 25 bp fragment or the authentic Yi element. A deletion of the 5'-flanking region of the 28 bp fragment up to 5 bases also abolished the binding activity. The CBP/tk binding in IMR-90 cells was supershifted by antiserum against NF-Ya, but not by antiserum made against p107, pRb, cyclin A, p33cdk2, or p34cdc2. Taken together, our results suggest that the G1/S regulatory cis-element in human TK promoter may be confined only to CBP/tk binding sites.

Tetrahymena ribozyme disrupts rRNA processing in yeast

The intervening sequence (IVS) of Tetrahymena thermophila nucleolar DNA interrupts a highly conserved sequence in the RNA core structure of the large ribosomal subunit. This location in nuclear DNA is unusual as most group I introns are in mitochondrial and chloroplast DNA. To examine the effect of a ribozyme insertion in another nuclear genome, the Tetrahymena IVS was introduced into the analogous position in a cloned Schizosaccharomyces pombe ribosomal gene, and the mutant rDNA was expressed in vivo. RNA analyses indicated that mature 5.8 S rRNA was not formed from the mutant gene transcript and the amount of 27 S nRNA was significantly reduced. In contrast, hybridization analyses indicated that RNA splicing continued, and normal forms of free ribozyme were present. The results show that the IVS sequence can interfere with rRNA processing and suggest that the unusual amplification of a single rDNA repeat may have forced Tetrahymena to accommodate its ribozyme.

Sagittal knee stability after anterior cruciate ligament reconstruction with a patellar tendon strip. A two-year follow-up study

Tibial anteroposterior displacement after anterior cruciate ligament reconstruction with a patellar tendon graft was followed prospectively for 2 years in 24 patients with an arthrometer. The femoral ligament insertion location, in a lateral projection, and the change in intraarticular fixation distance, measured with an isometer, were documented intraoperatively. Two years after surgery, the overall mean injured-noninjured difference in anteroposterior displacement was 2.0 +/- 2.3 mm. All grafts were fixed during surgery at 20 degrees of knee flexion. Patients for whom this angle coincided with the angle of minimum intraarticular fixation distance (Group I), and patients who had a femoral ligament insertion location > 2 mm anterior to the center of the normal anterior cruciate ligament attachment (Group A) showed larger tibial displacement than the other patients. An injured-noninjured difference in tibial anteroposterior displacement > or = 3 mm was classified as failure. Groups I and A failure rates were higher than for the other patients. No correlation was found between anteroposterior displacement and magnitude of the change in intraarticular fixation distance. We conclude that anterior femoral locations lead to larger sagittal play after 2 years than central or posterior locations and that the magnitude of the fixation distance is less important than the pattern.

Intragenic processing in yeast rRNA is dependent on the 3' external transcribed spacer

The nucleotide sequence of the 3' external transcribed spacer (3' ETS) region in Schizosaccharomyces pombe rDNA was determined to define structural features which mediate the termination of RNA transcription and subsequent rRNA maturation. S1 nuclease protection studies suggest three alternative termination sites and four cleavage sites in the processing of the 3' ETS sequence. Each of the termination sites precedes a "Sal box"-like sequence which has been demonstrated to mediate the termination of rRNA transcription in mammalian cells. A highly conserved extended hairpin structure in the ETS sequence was deleted by PCR-mediated mutagenesis and the mutant rDNA was expressed in vivo to determine its role in rRNA maturation. Despite an efficient expression of the mutant gene, mature 5.8 S or 25 S rRNA was not observed. Labelling kinetics and S1 nuclease protection analyses indicate that the deletion not only fully inhibits the removal of the 3' ETS but also fully inhibits the processive excision of the second internal transcribed spacer (ITS2). Instead, a relatively stable 27 S nRNA precursor remains easily detectable in the whole cell RNA population. The results demonstrate a critical dependence of ITS processing on the 3' ETS raising the possibility that these sequences interact in a common processing domain.