Uncovering and engineering the mechanical properties of the adhesion GPCR ADGRG1 GAIN domain.

Key cellular functions depend on the transduction of extracellular mechanical signals by specialized membrane receptors including adhesion G-protein coupled receptors (aGPCRs). While recently solved structures support aGPCR activation through shedding of the extracellular GAIN domain, the molecular mechanisms underpinning receptor mechanosensing remain poorly understood. When probed using single-molecule atomic force spectroscopy and molecular simulations, ADGRG1 GAIN dissociated from its tethered agonist at forces significantly higher than other reported signaling mechanoreceptors. Strong mechanical resistance was achieved through specific structural deformations and force propagation pathways under mechanical load. ADGRG1 GAIN variants computationally designed to lock the alpha and beta subdomains and rewire mechanically-induced structural deformations were found to modulate the GPS-Stachel rupture forces. Our study provides unprecedented insights into the molecular underpinnings of GAIN mechanical stability and paves the way for engineering mechanosensors, better understanding aGPCR function, and informing drug-discovery efforts targeting this important receptor class.

Anomalies of the TGF-beta postreceptor signaling pathway in ovarian cancer cell lines.

Transforming growth factor-beta (TGF-beta) can cause cell cycle arrest, terminal differentiation, or apoptosis in most normal epithelial cells, whereas most malignant cell lines are resistant to TGF-beta. Mechanisms of resistance to TGF-beta caused by modulation of cell cycle regulators and/or inactivation of components of the TGF-beta signaling transduction pathway such as C-myc and Smad4 have been demonstrated in human pancreatic cancer and squamous cell carcinoma cell lines. But, this has not been shown in ovarian cancer. To investigate the potential association between loss of sensitivity to TGF-beta and expression status of transforming growth factor receptor II (T beta RII), Smad4, CDC25A and C-myc in fourteen cell lines derived from ovarian cancer, the expression levels of these genes were examined by semi-quantitative RT-PCR. Normal ovarian surface tissues were used as controls. Expression of T beta RII was detectable in all of fourteen cell lines. Expression of Smad4 was decreased in ten cell lines and nine cell lines overexpressed CDC25A, compared to normal controls. CDC25A gene was overexpressed in 88% (8/9) of tumorigenic cell lines as determined by xenografts in nude mice, and only in 20% (1/5) of non-tumorigenic cell lines (P < 0.05). C-myc was not overexpressed in any of these cell lines. The loss of sensitivity to TGF-beta of cell lines derived from ovarian cancers may be related to (1) a decreased expression of Smad4, which mediates TGF-beta induced growth inhibition; and/or (2) an overexpression of CDC25A. This overexpression correlates with increased tumorigenicity of ovarian cancer cell lines. The loss of sensitivity to TGF-beta is not associated with a lack of T beta RII.

Expression of costimulatory molecules CD80 and CD86 and their receptors CD28, CTLA-4 on malignant ascites CD3+ tumour-infiltrating lymphocytes (TIL) from patients with ovarian and other types of peritoneal carcinomatosis.

Costimulation of T lymphocytes by the leucocyte surface molecules CD80 and CD86 expressed on antigen-presenting cells (APC) is required for the development of T cell responses. The CD28 and CTLA-4 molecules on T cells serve as receptors for the CD80 and CD86 costimulatory antigens. We have examined the frequency of expression of CD80 (B7.1), CD86 (B7.2), CD28 and CTLA-4 surface antigens on TIL isolated from malignant ascites or peritoneal washings of 26 patients with ovarian carcinoma and five patients with non-ovarian peritoneal carcinomatosis. Expression of CD80 and CD86 antigen was detected by reverse transcription-polymerase chain reaction (RT-PCR), and by FACS analysis. Significantly higher proportions of intraperitoneal CD3+ cells expressed CD86 antigen than the CD80 antigen (14 +/- 9% versus 3 +/- 3%, P < 0.05). Moreover, CD3+CD86+ cells were significantly more frequent in the peritoneal fluid (14 +/- 9%) than in the peripheral blood (3 +/- 0.4%, P < 0.05) of ovarian patients or normal controls (3 +/- 1%). CTLA-4 and CD28 antigen were expressed, respectively, on 9 +/- 4% and 86 +/- 14% of ascitic CD3+ cells of ovarian cancer patients. Both CD80 and CD86 antigens were expressed primarily on HLA-DR+ ascites TIL and were present in a very low proportion of HLA-DR- ascites TIL. These HLA-DR+ cells may represent a population of lymphocytes that have been activated in vivo, and function as APC. An anti-CD86 MoAb or a combination of anti-CD86 and anti-CD80 MoAbs significantly inhibited the proliferation of cultured intraperitoneal TIL. We have shown that in addition to CD28 and CTLA-4, CD3+ intraperitoneal TIL express the costimulatory molecules CD80 and CD86. The expression of these molecules on T cells could be dependent upon certain factors in the tumour microenvironment that could determine the outcome of in vivo immune responses.

Identification of an IL-10-producing HLA-DR-negative monocyte subset in the malignant ascites of patients with ovarian carcinoma that inhibits cytokine protein expression and proliferation of autologous T cells.

A newly described subset of monocytes has been identified in peritoneal exudate cells (PEC) from the malignant ascites from patients with ovarian cancer. These cells were characterized by the production of IL-10 and TGF-beta2, but not IL-12, IL-1alpha, or TNF-alpha, and they expressed CD14, CD16, and CD54, but not HLA-DR, CD80, CD86, CD11a, CD11b, or CD25 cell surface Ags. Since this subset of monocytes could affect the modulation of tumor immune responses in vivo, studies were undertaken to determine their effect on the activation and proliferation of autologous T cells from the peritoneal cavity of patients with ovarian carcinoma. Expression of cytokine-specific transcripts in T cells was determined by RT-PCR. Transcripts for the following cytokines were detected in patient specimens that also contained the IL-10-producing monocytes IL-2 (12 of 17 specimens), GM-CSF (9 of 17 specimens), IFN-gamma (6 of 17 specimens), and TNF-alpha (4 of 17 specimens). Cytokine production by T cells was determined by intracellular flow cytometry and by ELISA. IL-2 and IFN-gamma proteins, unlike their transcripts, were detected only in specimens that lacked IL-10-producing monocytes. IL-10-producing monocytes cocultured with autologous T cells inhibited the proliferation of the T cells in response to PHA. However, T cells cocultured with PEC from which the IL-10-producing monocytes had been removed did not inhibit T cell proliferation. Moreover, the inhibition of T cell proliferation by IL-10-producing monocytes could be reversed by adding neutralizing Abs to both IL-10R and TGF-beta2. These results suggest that this subset of monocytes may modulate immune responses by inhibiting T cell proliferation and cytokine protein production.

Quantitative analysis of transforming growth factor beta 1 and 2 in ovarian carcinoma.

Transforming growth factor beta (TGF-beta) is an important family of cytokines that may promote tumor growth in vivo through several mechanisms including interference with antitumor T-cell immune responses, alteration of factors in the stroma and matrix, and the promotion of angiogenesis. TGF-beta isotypes have been detected in malignant and normal ovarian tissues. We have determined by quantitative immunohistochemistry the density of TGF-beta1, TGF-beta2, and human leukocyte antigen (HLA) Class I and Class II antigens on malignant cells in paired primary and metastatic specimens from 10 patients with ovarian carcinoma. Cryostat sections of specimens from the carcinomas and from normal ovaries of three women of similar age without ovarian cancer were stained respectively with specific antibodies to TGF-beta1, TGF-beta2, and HLA Class I and II antigens, and with isotype-matched control antibodies. Antigen density was quantitated blindly as mean absorbance on a SAMBA 4000 image analyzer. TGF-beta1 and TGF-beta2 were overexpressed in both primary and metastatic tumor specimens in comparison with normal ovarian tissue. No statistical correlation was found between the expression of TGF-beta1 or TGF-beta2 and HLA class I or HLA class II, which suggests that TGF-beta isotypes could have effects on the immune system other than down-modulation of these HLA molecules. Furthermore, the lack of association between levels of TGF-beta expression and the reduced expression of HLA molecules could suggest that tumor cells expressing both HLA and TGF-beta may be suitable targets for adaptive immunotherapy. Additional studies are necessary to determine whether TGF-beta expressed by ovarian cancer cells merits evaluation as a therapeutic target.

In vitro growth inhibition of ovarian cancer cells by decorin: synergism of action between decorin and carboplatin.

In vitro studies showed that decorin, a small proteoglycan that is a normal component of the cell matrix involved in tissue scaffolding, effectively inhibited the growth of two ovarian cancer lines, SKOV3 and 2774. Using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay to measure cell growth, IC50s for decorin ranged from 150 to 400 microg/ml for the two cell lines. In contrast, the growth of tumor cells grown on an artificial cell matrix (Matrigel) was unaffected by decorin treatment, perhaps because of the decorin being irreversibly bound by matrix-associated collagen. Decorin-induced inhibition of ovarian tumor cells appeared to be associated with the increased expression of the cyclin-dependent kinase inhibitor p21Waf1/Cip1. Up-regulation of p21 expression was shown by Western blot analysis in decorin-treated ovarian cancer cells. No decorin-induced up-regulation of c-myc was seen, although decorin was reported to activate the epidermal growth factor receptor. Decorin was also shown to synergize with carboplatin to inhibit the growth of ovarian tumor cells. Additional studies are warranted to determine the role of decorin in the treatment of ovarian cancer.

The role of cytokines in both the normal and malignant ovary.

Normal ovarian tissue is rich in cytokines. Cytokines and chemokines are important in the physiology of ovarian function and of ovulation. Cytokines and chemokines may recruit cytokine-producing lymphocytes to the site of a developing follicle, and cytokines appear to play an important role in pre and post follicle development. Most of the same cytokines that are found in normal ovarian tissue are also found in association with malignancy in contrast to their functions in normal tissues. It is reasonable to assume that the functions of cytokines associated with malignancy may serve to promote the unregulated growth if tumor cells and metastasis. It is also likely that cytokines produced by tumors will modulate immune responses that favor tumor progression. In the following review, we have highlighted those functions of cytokines that have been identified as having the most significant impact on tumor growth and development. By examining activities of these cytokines in normal and in malignant ovarian tissues, it is hoped that future possible avenues for investigation may be opened up and that the results of these investigations will lead to strategies that can modulate the production or the activity of the cytokines leading to the growth of tumors or their metastases. Such strategies now fall under the general discipline of bioimmunotherapy. This is an expanding discipline as more is learned about growth regulation in cancer, and with the availability and rapid development of new molecules for therapeutic approaches.

RT-PCR quantitation of cytokine responses in vivo from specimens containing small numbers of cells during bioimmunotherapy.

A Standard Template method has been developed to carry out semiquantitative RT-PCR analysis on mRNA extracted from small specimens that contain low yields of RNA. Using easily prepared templates (made from previously tested reference specimens), standard calibration curves were generated for each of two cytokine products of interest, specifically IL-10 and IFN-gamma. Also, internal standardization was accomplished by quantitating a well-characterized housekeeping gene (GAPDH). Simple densitometry of the RT-PCR product did not demonstrate sufficient reliability for quantitation since a logarithmic relationship was shown between product and template input. Peritoneal exudate cell specimens that were obtained from ovarian cancer patients during intraperitoneal immunotherapy were utilized for the demonstration of IL-10 and IFN-gamma transcript in vivo. Briefly, this method consists of: (1) template preparation: a PCR product for the cytokine of interest is generated from a previously identified positive specimen, purified and stored at -20 degrees C. (2) RT-PCR: cDNAs are produced from RNA extracted from patient specimens. Replicates of each cDNA and serial dilutions of the corresponding template are amplified with primers specific for each cytokine of interest. (3) Quantitation: photographs are made of the PCR products displayed on an agarose gel and are analyzed by densitometry. Determinations of fold-differences in cytokine transcript expression are normalized to the mRNA content of each specimen (based on the expression of GAPDH).

Differential expression of cytokine transcripts in human epithelial ovarian carcinoma by solid tumour specimens, peritoneal exudate cells containing tumour, tumour-infiltrating lymphocyte (TIL)-derived T cell lines and established tumour cell lines.

T cell lines derived in low concentrations of recombinant IL-2 (rIL-2) from TIL of patients with epithelial ovarian carcinoma (EOC) often exhibit specific cytotoxicity against autologous tumour cells. However, the ability of T cells at the tumour site to respond to ovarian carcinoma cells may be affected by the production of cytokines by the various cell types present. Using reverse transcriptase-polymerase chain reaction (RT-PCR) we investigated cytokine transcripts in: (i) established EOC tumour cell lines; (ii) solid tumour specimens or peritoneal exudate cells (PEC) from ascites or peritoneal washings of patients with EOC; and (iii) CD4+ TCRalphabeta+ and CD8+ TCRalphabeta+ TIL-derived T cell lines developed in rIL-2. We have found that (i) established EOC tumour cell lines expressed transcripts for transforming growth factor-beta 2 (TGF-beta2) (7/7), but not IL-10 (0/7) or interferon-gamma (IFN-gamma) (0/7) and rarely IL-2 (1/7); (ii) PEC expressed transcripts for IL-2 (12/13), IL-10 (9/13), and TGF-beta2 (12/13), and less often, IFN-gamma (3/13), whereas solid tumour specimens from eight patients with EOC expressed transcripts for IL-2 (4/8), TGF-beta2 (4/8), and IL-10 (5/8), but not for IFN-gamma (0/8); (iii) CD4+ TCRalphabeta+ T cell lines expressed transcripts for IFN-gamma (4/4), IL-2 (4/4) and IL-10 (3/4), whereas CD8+ TCRalphabeta+ T cell lines expressed transcripts for IFN-gamma (5/5), IL-2 (1/5) and IL-10 (2/5). None of these T cell lines expressed TGF-beta2 transcripts. The frequency of IL-2 and TGF-beta2 transcripts in solid tumours was significantly lower than in the PEC (P = 0.0475). CD4+ or CD8+ T cell lines expressing IFN-gamma, IL-2 and IL-10 transcripts were derived in culture with rIL-2 from the TIL of specimens that did not necessarily express these cytokines in the absence of rIL-2. The frequency of cytokine transcripts in T cell lines compared with these same transcripts in the PEC was significantly higher for IFN-gamma (P = 0.0005) and lower for TGF-beta2 (P = 0.0001). An association was observed between the expression of cytokine transcripts in vivo or by TIL-derived cell lines and functions exhibited by either production of cytokines or in vitro cytotoxicity.

Immunopharmacology and cytokine production of a low-dose schedule of intraperitoneally administered human recombinant interleukin-2 in patients with advanced epithelial ovarian carcinoma.

We determined in the peritoneal cavity (p.c.) of epithelial ovarian carcinoma patients during a 4-day treatment cycle of low-dose recombinant human interleukin-2 (rIL-2): (a) pharmacokinetics of IL-2, (b) endogenous cytokine production, and (c) numbers and percentages of peritoneal exudate lymphocytes. We administered 6 x 10(5) IU/m2 of rIL-2 (0.03 mg/m2 Proleukin rIL-2) intraperitoneally (i.p.) over 30 min on each of 4 days. One and one-half liters of D5 0.25 NS was injected i.p. before each rIL-2 infusion. Multiple peritoneal fluid samples were obtained from each of four patients on day 1 and day 4 for detection of IL-2, endogenous cytokines, and soluble IL-2 receptor (IL-2R-alpha). IL-2 concentrations in the peritoneal fluid were determined by bioassay and interferon (IFN)-gamma, tumor necrosis factor (TNF)-alpha, IL-10, transforming growth factor (TGF)-beta 2, and sIL-2R-alpha by enzyme-linked immunosorbent assay. Numbers of cells per microliter and lymphocyte subpopulation percentages after staining with a panel of monoclonal antibodies were determined on day 1, day 4, and subsequent off-treatment days. IL-2 disappearance in the p.c. was well described by a pharmacokinetic model having constant-rate infusion and biexponential disposition. About 90% of the IL-2 disappearance occurred during the beta-phase, during which IL-2 concentrations were sustained at approximately 10-30 ng/ml (day 1 and day 4) and the median t1/2 beta was 21.5 and 9.2 h on days 1 and 4, respectively. In four of four patients, p.c. production of IL-10 was observed on day 1 and day 4 (maximum 387 pg/ml). Maximum levels of IFN-gamma and sIL-2R-alpha were observed on day 4. (IFN-gamma 217 pg/ml; sIL2-R-alpha: 3486 U/ml). No increases in TNF-alpha or TGF-beta 2 were observed. Large increases in p.c. CD3+, CD4+, CD8+, CD16+, and CD56+ cells were observed. We conclude that biologically active levels of IL-2 are generated in p.c. fluids after i.p. administration of rIL-2 at 0.03 mg/m2.

Transduction of rIL-2 expanded CD4+ and CD8+ ovarian TIL-derived T cell lines with the G1Na (neor) replication-deficient retroviral vector.

We have expanded ovarian tumor-infiltrating lymphocytes (TIL) in low concentrations of recombinant interleukin-2 (rIL-2) to conduct intraperitoneal adoptive immunotherapy trials in patients with ovarian cancer. We have previously demonstrated that certain T cell lines and clones derived from ovarian TIL exhibit in vitro autologous tumor-specific cytotoxicity and/or cytokine production (interferon-gamma, tumor necrosis factor-alpha) preferentially in response to autologous tumor cells. Studies that utilize a marker gene introduced into the DNA of TIL can provide useful information on specific uptake or localization of TIL at tumor sites and on the survival of TIL in vivo. We have conducted a series of preclinical experiments in which we have successfully transfected TIL with G1Na, which encodes the gene for neomycin phosphotransferase (neoR). NeoR was detected in at least 10% of CD8+ cells (mean = 10.4%) and between 2.5 and 20% of CD4+ TIL (mean = 8.5%). Transduction of ovarian TIL with G1Na caused no substantial changes to the T cell phenotypes or in vitro cytotoxicities against ovarian and hematogenous tumor cell targets, or on the rIL-2 requirements of TIL for growth and proliferation. In addition, the intact G1Na provirus in transduced TIL cells was rescuable by replication-competent retrovirus and was transferred into the genome of NIH-3T3 fibroblasts, which were rendered resistant to G418. An enhanced polymerase chain reaction (PCR) procedure utilizing detection by ethidium bromide staining was developed. The enhanced PCR detected 1 in 100,000 neoR-labeled cells. Furthermore, detection of the G1Na genome in transduced TIL by in situ hybridization with an RNA probe provided evidence for expression of the neoR gene in transduced TIL. Results obtained from these studies suggest that ovarian TIL-derived T cell lines transduced with the neoR gene post infection with the G1Na retroviral vector can be utilized to examine the in vivo trafficking pattern of ovarian TIL-derived T cell lines expanded in low concentrations of rIL-2 and their survival.

Identification of a new viral protein containing CAp30 and NCp10 sequences in murine and feline leukemia retroviruses.

Because Pr65gag is in part located in the nucleus and contains a putative bipartite nuclear targeting signal, we investigated the cellular location and structure of P55gag, a gag-encoded polyprotein known to lack the nucleocapsid (NC) protein NCp10. P55gag was found to be restricted to the cytoplasm of Moloney murine leukemia virus-infected cells. Of interest, P55gag was produced in cells infected by a viral protease deletion mutant and by a recombinant murine sarcoma virus known to lack the protease gene. Surprisingly, our structural and immunological studies indicated that P55gag also lacks carboxy-terminal residues of CAp30. During the course of studying P55gag, we detected a new viral protein within purified virus particles that contained NCp10 tryptic peptide sequences and a CAp30 tryptic peptide lacking in P55gag. This viral protein, which we have named nucleocapsid-related protein (NCRP), also contained antigenic epitopes present in CAp30 and NCp10. P55gag- and NCRP-like proteins were also observed in AKV murine leukemia virus and feline leukemia virus systems. The precise site of cleavage within Pr65gag that produces P55gag and NCRP is unknown but lies upstream of the CAp30-NCp10 junction within the carboxy-terminal domain of CAp30. The existence of a form of NCp10 containing carboxy-terminal CAp30 sequences raises interesting possibilities about its functional role in genomic RNA packaging and/or viral RNA dimerization.

A subset of Pr65gag is nucleus associated in murine leukemia virus-infected cells.

Nuclei of cells infected with Moloney murine leukemia virus (MoMuLV) were examined for the presence of gag proteins. This analysis was performed in conjunction with other studies suggesting a possible role for gag proteins in regulating nuclear events relating to processing and/or transport of viral genomic RNA. We detected Pr65gag and a p30-related protein in a nuclear fraction of infected cells. We also found evidence that a highly conserved amino acid sequence, which is shared by p30 and U1 small nuclear ribonucleoprotein 70-kDa protein, is a component of the nuclear targeting sequence for Pr65gag. Immunoelectron microscopy studies with a monoclonal anti-p12 antibody established that approximately 18% of gag-containing proteins of MoMuLV are located in the nucleus. Such gag-containing proteins from a mutant MoMuLV that lacks N-terminal myristic acid had greater affinity for the nucleus, suggesting that fatty acid acylation of Pr65gag plays a role in overcoming the proposed nuclear transport signal. The possible roles that nuclear gag proteins may play in retroviral replication are discussed.

Novel GACG-hairpin pair motif in the 5' untranslated region of type C retroviruses related to murine leukemia virus.

We searched for the presence of common RNA structural motifs in mammalian type C retroviruses related to murine leukemia viruses and the closely related avian spleen necrosis virus. A novel motif consisting of a pair of hairpins, called hairpin pair motif, was detected in the 5' untranslated regions of the genomes of these retroviruses. A combination of computational analyses that included the assessment of phylogenetic sequence conservation by multiple alignment, the search for regions with unusual RNA folding properties, and the analysis of RNA secondary structure by suboptimal free-energy calculations highlighted the significance of this hairpin pair motif. The hairpin pair motif encompasses 70 to 80 nucleotides between the splice donor site and the gag translational initiation codon of these viruses. The motif is composed of two adjacent hairpins both with a perfectly conserved GACG tetraloop. We propose that the novel GACG-hairpin pair motif described here constitutes an essential component of the regulatory machinery in these type C retroviruses.

A developmental switch in sea urchin U1 RNA.

The sequence of U1 RNA has been determined in the eggs and embryos of two sea urchins, Lytechinus variegatus and Strongylocentrotus purpuratus. In both species the sequence of the U1 RNA changes as the embryos progress through development. The sequence of the major U1 RNA in the eggs of the two species differs in two nucleotides, while the sequence of the U1 RNA present in the late embryos and somatic tissue is identical in the two species. The U1 RNA in eggs and early embryos is primarily derived from the tandemly repeated gene set, which is not expressed in somatic tissues.

Structure of an unusual sea urchin U1 RNA gene cluster.

Genomic clones containing multiple copies of the Lytechinus variegatus U1 gene have been isolated from a gene library in the phage lambda EMBL3. These clones contain both types of U1 RNA gene repeats interspersed in the same 15-kb fragment. In addition, about 1/3 of the repeat units contain a 260-bp insert 460 bp prior to the first nucleotide of the U1 RNA sequence. The inserted sequence is abundant in the sea urchin genome as judged by Southern blots of genomic DNA. There are no repeated sequences flanking the insert. The insert occurs at the same position in the highly conserved 5'-flanking region at which a deletion has previously been reported.

Sea urchin maternal and embryonic U1 RNAs are spatially segregated in early embryos.

We have used in situ hybridization and cell fractionation methods to follow the distribution of U1 RNA and immunofluorescence microscopy to follow the distribution of snRNP proteins in oocytes, eggs, and embryos of several sea urchin species. U1 RNA and U1-specific snRNP antigens are concentrated in germinal vesicles of oocytes. Both appear to relocate after oocyte maturation because they are found primarily, if not exclusively, in the cytoplasm of mature unfertilized eggs. This cytoplasmic residence is maintained during early cleavage and U1 RNA is first detectable in nuclei of micromeres at the 16-cell stage. Between morula and gastrula stages the steady-state concentrations of both RNA and antigens gradually increase in nuclei and decrease in cytoplasm. Surprisingly, analysis of the distribution of newly synthesized U1 RNA shows that it does not equilibrate with the maternal pool. Instead new transcripts are confined to nuclei, while cytoplasmic U1 RNAs are of maternal origin. This lack of equilibration and the conversion of maternal U1 RNAs from nuclear species in oocytes to cytoplasmic in embryos suggests that these RNPs (or RNAs) are structurally altered when released to the cytoplasm at oocyte maturation.

Structure and expression of a second sea urchin U1 RNA gene repeat.

There are two tandemly repeated sets of U1 RNA genes in the sea urchin L. variegatus. Each of these genes is present in a 1.4 kb repeat defined by a HindIII site about 450 bases 5' to the gene. The sequences of a member of both repeating units (U1.1 and U1.2) has been determined. The repeats are nearly identical for 550 nucleotides 5' to the gene but show great divergence starting 30 nucleotides 3' to the gene, just after the CAAAGAAAGAAAA sequence thought to be required for 3' end formation. The other boundary between the conserved and non-conserved sequences is a polypyrimidine sequence (on the strand which codes for U1 RNA). Both of these repeats are expressed in blastula stage embryos, as judged by transcription of unique sequences 3' to the gene in nuclei isolated from blastula stage embryos. At least some of the two types of repeats are interspersed, since representatives of both repeat types on a single gamma phage isolated from a gene library. The sequence of the U1 RNA in L. variegatus eggs and embryos corresponds to the sequence of the U1 repeat.