Search

In order to analyze the RNA populations present in different cells of very early embryos, we have developed a protocol to purify these large blastomeres using counterflow centrifugal elutriation (CCE). This procedure employs ethanol fixation to stabilize the cells against shear forces encountered during CCE. Using this method, we fractionated the three different blastomere types of the 16-cell sea urchin embryo, the micromeres, mesomeres, and macromeres, achieving 96, 94, and 96% mean purities, respectively. We show here that intact RNA is recovered with equal efficiency from each blastomere preparation. Using this method, we have identified several RNAs that are distributed non-uniformly among these cells.

We have identified an mRNA that encodes a protein, SpS24, of the small ribosomal subunit in the sea urchin, Strongylocentrotus purpuratus. RNA blot and in situ hybridization analyses show that the SpS24 gene is active during early oogenesis, downregulated in the mature egg and during cleavage, and reactivated in the early blastula. The mRNA then increases in abundance at least 100-fold. Later in development, expression of SpS24 mRNA becomes restricted primarily to cells in the oral ectoderm and endoderm of the pluteus larva, and the message is undetectable in aboral ectoderm cells and most mesenchyme cells. To determine whether transcription of the ribosomal RNA genes occurs at a higher rate in oral ectoderm and endoderm tissues, a probe for the transcribed spacer was used in RNase protection and in situ hybridization assays. High concentrations of rRNA-processing intermediates were observed in unfertilized eggs and shown to reside primarily, if not exclusively, in the cytoplasm. The spatial and temporal distributions of these sequences strongly suggest that they are associated with heavy bodies. New embryonic rRNA transcripts are first detectable at the very early blastula stage. In later embryos, the content of this transcribed spacer sequence is similar in all but a few cells, which implies that they synthesize rRNA at a similar low rate. Comparison of available estimates of rRNA transcription rate with the potential rate of SpS24 protein synthesis, calculated from SpS24 mRNA prevalence, shows that oral ectoderm and endoderm cells have the capacity to synthesize 15- to 30-fold more SpS24 protein than is required to keep pace with rRNA synthesis in these cells. Because the sea urchin embryo develops from an egg to a pluteus larva in the absence of growth, this stockpiling of SpS24 mRNA anticipates rather than accompanies the onset of growth, which does not begin until after feeding. Upregulation of this gene is therefore part of the developmental program, rather than a physiological response to nutrient availability.

Two major signaling centers have been shown to control patterning of sea urchin embryos. Canonical Wnt signaling in vegetal blastomeres and Nodal signaling in presumptive oral ectoderm are necessary and sufficient to initiate patterning along the primary and secondary axes, respectively. Here we define and characterize a third patterning center, the animal pole domain (APD), which contains neurogenic ectoderm, and can oppose Wnt and Nodal signaling. The regulatory influence of the APD is normally restricted to the animal pole region, but can operate in most cells of the embryo because, in the absence of Wnt and Nodal, the APD expands throughout the embryo. We have identified many constituent APD regulatory genes expressed in the early blastula and have shown that expression of most of them requires Six3 function. Furthermore, Six3 is necessary for the differentiation of diverse cell types in the APD, including the neurogenic animal plate and immediately flanking ectoderm, indicating that it functions at or near the top of several APD gene regulatory networks. Remarkably, it is also sufficient to respecify the fates of cells in the rest of the embryo, generating an embryo consisting of a greatly expanded, but correctly patterned, APD. A fraction of the large group of Six3-dependent regulatory proteins are orthologous to those expressed in the vertebrate forebrain, suggesting that they controlled formation of the early neurogenic domain in the common deuterostome ancestor of echinoderms and vertebrates.

Publisher Summary This chapter discusses the localization of messenger RNAs (mRNAs) by in situ hybridization. In situ hybridization helps in distinguishing the cells in a complex tissue that express specific mRNAs and in making semiquantitative estimates of the relative concentration of these in different cell types. Other methods of analysis of RNA isolated from cells, tissues, or organisms—such as RNA blots or solution hybridization—average the content of a specific mRNA over all cell types in the sample. In situ hybridization offers several other potential advantages over solution or filter hybridization methods. The aim and many of the advantages of in situ hybridization are shared with the immunocytochemical detection of the encoded proteins. The basic difference is that they monitor gene expression at different levels. By measuring mRNA concentrations, a more accurate estimate of gene activity is provided because mRNA accumulates before the protein it encodes, and because relatively stable proteins may persist long after the gene is repressed and the mRNA has decayed. The technique described in the chapter employs the hybridization of radiolabeled riboprobes to sections of aldehyde-fixed and paraffin-embedded tissue.

Spec 1 mRNAs increase 100-fold in abundance per embryo during early sea urchin development. Previous studies indicated an enrichment of this mRNA in ectoderm fractions of gastrulae and plutei. We have determined the precise localization of this mRNA by in situ hybridization techniques. In pluteus larvae, the mRNA is highly restricted to a set of morphologically uniform ectoderm cells in the dorsal part of the embryo. The mRNA is not detectable in other regions of ectoderm or in endoderm and mesoderm. The pattern of localization is already established at the gastrula stage, before these cells are distinguishable by morphological criteria. This pattern of distribution of Spec 1 mRNA is distinct from that of bulk poly(A)+ mRNA. Measurements of the amount of Spec 1 mRNA per embryo and the number of cells containing this RNA indicate that there are about 500 Spec 1 mRNA molecules per cell at the pluteus stage and probably twice as many at the gastrula stage. These results indicate that the sensitivity of the in situ hybridization method allows detection of sequences that comprise approximately equal to 0.05% of the embryo mRNA nucleotides.

The transition from early (E) to late (L) histone gene expression in developing sea urchin (Strongylocentrotus purpuratus) embryos was examined for H2B, H3, and H4 mRNAs by in situ hybridization of class-specific probes. Hybridization patterns indicate that the shift from E to L mRNAs occurs gradually and simultaneously in all blastomeres. Thus, during the transition the ratio of L to E mRNAs is similar in most cells. This suggests that no sudden changes in histone composition occur in individual cells which might be related to alterations in gene expression associated with differentiation of cell lineages. Around the midpoint of the transition, clusters of cells progressively appear which contain little, if any, E or L histone mRNA. This modulation of expression is coordinated for the three late genes examined because most individual cells contain either high or low levels of all three mRNAs. At blastula stage these clusters of unlabeled cells appear to be randomly distributed throughout the embryo. Subsequently the unlabeled regions expand and are found predominantly in aboral ectoderm as these cells cease to divide. Thus, the L E histone mRNA ratio is not differentially regulated in diverse cell lineages, and the major differences in total histone mRNA content among individual cells may be related to cell cycle and/or the cessation of division.

The calculated rate of protein synthesis for growing Tetrahymena is 360 pg/h, whereas starved cells synthesize only about 3 pg of protein/h. Within 50 min after deciliation of starved cells, the rate of protein synthesis increases to about 60 pg/h. The major mechanism to accomplish these large and rapid changes in the rate of bulk protein synthesis involves regulation of the number of messages loaded on polysomes. Logarithmically growing cells contain about 3.2 X 10(7) mRNA molecules/cell, of which approximately 60% is loaded on polysomes. Starved cells contain about 0.8 X 10(7) messages and the percentage of messages loaded is reduced to 4%. Thus, the number of loaded messages is approximately 60-fold lower in starved cells than in growing cells, although the total message content of cells in these two physiological states differs by only a factor of 4. After deciliation of starved cells, message loading increases about 10-fold. It seems likely that much of the message loaded after deciliation is derived from the large pool of nonpolysomal message in starved cells. Although large differences in message loading exist for growing, starved, and starved-deciliated cells, measurements of the rate of polypeptide elongation and the rate of message initiation indicate the translational efficiency of loaded messages (pg of protein synthesized per pg of message/unit time) is very similar under all conditions.

Members of the Spec gene family are expressed during embryonic development of the sea urchin, Strongylocentrotus purpuratus. The family encodes proteins related to the calmodulin/troponin C/myosin light chain group of calcium binding proteins and one gene, Spec1, has been studied extensively in our laboratory. In this paper, we analyze other members of the family, collectively termed Spec2 genes. We make use of several hybridization probes derived from Spec1 and Spec2 cDNA clones, which recognize different members of the family. Genomic DNA gel blot and slot blot analyses show that there are approximately eight Spec genes in the S. purpuratus genome. The structures of three Spec2 genes, Spec2a, Spec2c and Spec2d, are described. A 60 kb (kb = 10(3) bases or base-pairs) region of the genome contains the linked Spec1-Spec2c genes and two separate 20 kb regions contain the Spec2a and Spec2d genes. Six members of a repetitive sequence family are dispersed at various locations among the genes. The transcriptional initiation sites of the three Spec2 genes are mapped, and 400 to 500 base-pairs of 5'-flanking DNA sequenced. All three Spec2 genes initiate transcription approximately 120 base-pairs upstream from the 3' end of the first exon. In contrast, the 5' end of the Spec1 transcript begins about 107 base-pairs farther upstream, so it contains 5' untranslated sequences that correspond to non-transcribed 5'-flanking sequences of the Spec2 genes. There is little similarity among the sequences upstream from the CAP site of the Spec2 genes except the TATA consensus sequence and a repeating trinucleotide, AAC. Measurements of Spec mRNA levels during embryogenesis show that Spec1 mRNA begins to accumulate at the early blastula stage and is the most abundant; Spec2a/Spec2c mRNAs begin accumulating several hours later at the late blastula-early gastrula stage and reach about 40 to 60% the levels of Spec1; and Spec2d mRNAs accumulate mostly during the gastrula and pluteus stages with levels reaching only 2% those of Spec1. In situ hybridization with probes that recognize either all Spec2 mRNAs or only Spec2d mRNAs show that, like Spec1, these mRNAs are restricted to aboral ectoderm cells and their precursors. The Spec gene family represents a group of related genes whose mRNAs all accumulate in the same cell type but at different times and to different levels during embryogenesis.

Measurements of the divergence of single copy DNA sequences among four sea urichin species are presented. At a standard criterion for reassociation (0.12 M phosphate buffer, 60 degrees C, hydroxyapatite binding) we observe the following extents of reaction and reductions in thermal stability for single copy DNA reassociation between Strongylocentrotus purpuratus tracer and heterologous driver DNA: S. dröbachiensis 68% and 2.5 degrees C; S. franciscanus 51% and 3.5 degrees C; Lytechinus pictus 12% and 7.5 degrees C. The implied extents of sequence relatedness are consistent with the phylogenetic relationships of these species. The rate of single copy sequence divergence in the evolutionary lines leading to the Strongylocentrotus species is estimated to be 0.06-0.35% per million years. The rate of divergence of total single copy sequence has been compared to that of structural gene sequences represented in S. purpuratus gastrula polysomal messenger RNA. When closely related species, S. purpuratus and S. franciscanus, are compared, these polysomal sequences are found to diverge at a lower rate than does the total single copy sequence. For two very distantly related species, S. purpuratus and L. pictus, a small fraction of the single copy DNA sequence is probably conserved. These conserved sequences are not enriched in their content of structural gene sequences.

Metallothioneins (MTs) are small, cysteine-rich proteins that bind heavy metals which induce their synthesis. Tissue fractionation of embryos at pluteus stage previously demonstrated that in the absence of added zinc, basal expression of MT mRNA is confined to ectoderm, whereas induction by zinc results in increased expression in the endoderm + mesoderm tissue fraction. Using in situ hybridization we now show that expression in the pluteus larva is restricted almost exclusively to the single cell type comprising the aboral ectoderm. Induction by Zn results in a marked accumulation of MT mRNA in gut and oral ectoderm to levels at least as high as that in aboral ectoderm. MT mRNA is also expressed in presumptive aboral ectoderm at earlier stages of normal development. In addition it is transiently expressed at variable levels in oral ectoderm and, to a lesser extent, in presumptive gut.

The number of molecules of mRNA transcribed from each of five different actin genes are reported for developing embryos of the sea urchin Strongylocentrotus purpuratus. Transcripts of the cytoskeletal actin genes CyI, CyIIa, CyIIb and CyIIIa, and of the muscle actin gene M, were measured in unfertilized egg and embryo RNAs of cleavage, blastula, gastrula and pluteus stages. The measurements were obtained by probe excess titrations of these RNAs, using a set of single-stranded RNA probes each identifying the mRNA transcripts of a specific actin gene. These mRNAs can be identified by their distinct 3′ non-translated trailer sequences. We confirm prior observations that the prevalence of actin mRNA in the unfertilized egg is low. Cytoskeletal actin genes CyI and CyIIIa each contribute 1 × 10^3 to 2 × 10^3 maternal mRNA molecules, and CyIIb contributes < 2 × 10^2 mRNA molecules, while no detectable maternal mRNAs derive from cytoskeletal actin gene CyIIa or the muscle actin gene M. During certain periods of development, transcripts derived from the individual cytoskeletal actin genes accumulate rapidly, with kinetics specific to each mRNA. Transcripts of the muscle actin gene are absent until after gastrulation, when the initial muscle progenitor cells are formed. At late stages of development, each of the five genes studied is represented by 10^4 to 10^5 mRNA molecules per embryo. The present measurements permit calculation of the levels of each actin mRNA species in the particular cell types in which each gene functions in the fully differentiated embryo.

Typical histone genes lack intervening sequences and encode small mRNAs (400-800 nucleotides) with short leader and trailer regions. Most histone mRNAs are not polyadenylylated but rather terminate in a highly conserved stem and loop structure. The early, late, and testis-specific histone genes of sea urchins, described to date, have this typical histone gene structure. We have identified an unusual H1 gene, H1-delta, in sea urchins that encodes a poly(A)+ mRNA. This mRNA is one of a group of polyadenylylated transcripts homologous with H1 gene probes. The sequence of H1-delta had been determined. H1-delta encodes a different H1 protein. Although the temporal expression of H1-delta mRNA is similar to that of other late H1 (beta and gamma) mRNAs, its spatial distribution at the time of maximal accumulation is distinct and confirms that H1-delta is regulated differently than other H1 genes.

We have previously described cloning of an exon of a sea urchin collagen gene and shown that its expression is temporally regulated during embryogenesis, beginning during blastula formation. We have now localized the protein encoded by the gene and the sites of its mRNA synthesis in the developing embryo. Antibody to a synthetic peptide reacts with a 208,000 Mr protein that is digestible by collagenase. Fractionation of pluteus stage embryos demonstrates that the protein is localized primarily with cells that form the syncytium of primary mesenchyme that elaborates the larval endoskeleton; furthermore, immunofluorescence localizes the epitope to the periphery of the endoskeleton in situ. Transcripts of the gene accumulate only in mesenchyme cells, especially those of the primary mesenchyme lineage. Measurements of absolute transcript abundance show that collagen mRNA is present in blastula primary mesenchyme cells at 600-700 copies per cell and at about fourfold lower amounts in other mesenchyme cells.

Asymmetric RNA probes, which contain only the mRNA coding strand, provide a large increase in hybridization efficiency in situ over that observed with either symmetric (both strands represented) RNA or DNA probes. Asymmetric RNA probes are synthesized in vitro by transcription from recombinants formed between sequences encoding sea urchin mRNAs and the transcription vector R7 delta 7. Using a probe representing early variant histone mRNA sequences we have characterized hybridization to sections of sea urchin embryos with respect to thermal stability of the hybrids formed, optimum temperature, effect of sequence divergence on hybrid thermal stability, and dependence of the hybridization signals on probe concentration and hybridization time. Estimates from the observed signals indicate that a large fraction of target RNAs is both retained in sections and hybridized with probe at saturation. Coupled with measurements of nonspecific background binding of heterologous probes, these data indicate that the method has sufficient sensitivity to detect many moderately abundant mRNAs (20-75 molecules per cell in the 1500-cell pluteus). In situ hybridizations to embryos at different developmental stages show that while histone mRNAs are uniformly distributed in cleaving embryos, different cell lineages of older embryos show large differences in accumulation of these mRNAs.

The complexity of messenger RNA in growing or starved Tetrahymena thermophila is similar and unusually high (approximately 4.5 X 10(7) nucleotides). The complexity of nuclear RNA in growing cells (approximately 7.8 X 10(7) nucleotides) is only about 1.7 times that of mRNA. The concentration of complex class (rare) messages (approximately 53 copies/growing cell and approximately 11 copies/starved cell) is low in comparison to the size of the cell. The concentration of complex nuclear transcripts is also very low (approximately 0.7 copies/growing cell nucleus and approximately 2.6 copies/starved cell nucleus) considering that the macronucleus contains 45 to 90 copies of each single copy sequence. The complex sequence sets found on polysomes of growing and starved cells overlap about 80% and about 60% of the complex nuclear transcripts appear to be held in common. About 60% of macronuclear single copy DNA is transcribed in one or both physiological states. Although growing and starved cells have extremely different fractions of their messages loaded onto polysomes, within each cell type the complex messages in polysomal and nonpolysomal cytoplasmic fractions are indistinguishable, suggesting that exchange may occur between loaded and unloaded messages. Although T. thermophila DNA has an unusually low G + C content (23%), sequences coding for complex RNAs have base ratios similar to those of total DNA. Therefore, codon usage in Tetrahymena must be extremely biased towards adenine- and uridine-rich codons.

A homeo box-containing gene, Hbox1 is expressed in an unusual and highly conserved spatial pattern in embryos of two different species of sea urchin, Tripneustes gratilla and Strongylocentrotus purpuratus. Hybridization in situ shows that this mRNA accumulates initially throughout the aboral ectoderm; however, between blastula and pluteus stages, the region containing Hbox1 mRNA retracts gradually until only a small area around the vertex is labeled in pluteus larvae. Aboral ectoderm appears cytologically uniform and also accumulates uniform levels of other tissue-specific mRNAs. Therefore, the Hbox1 pattern reveals a previously unsuspected heterogeneity of aboral ectoderm cells and a polarity within this tissue. In S. purpuratus, the Hbox1 gene product probably is not involved in initial specification of cell fate, as this message does not achieve a significant fraction of its peak abundance until almost hatching blastula stage, well after the time aboral ectoderm cells have initiated a tissue-specific program of gene expression. RNA blot and RNase protection analyses revealed low levels of Hbox1 mRNA in all adult tissues examined. However, this message was not detectable in mature eggs, suggesting that the Hbox1 gene does not have a maternal function. In addition to highly conserved spatial and temporal patterns of expression, the homeo box genes of these two urchin species also are conserved highly in sequences outside the homeo domain, despite the divergence of these two species (30-45 my). Two notable features of the protein shared with several vertebrate homeo proteins are a short conserved sequence encoded by an exon upstream of that encoding the homeo domain and a large region of high serine and proline content.

We have used in situ hybridization and RNA blotting analysis to compare the timing of accumulation of poly(A) and α-subtype histone mRNA during oogenesis in the sea urchin Strongylocentrotus purpuratus . In situ hybridization with 3 H-poly(U) shows that the content of poly(A) in the developing oocyte increases four- to sixfold during vitellogenesis, implying a similar increase for polyadenylated maternal RNAs. In contrast, both RNA blotting and in situ hybridization demonstrate that there is little, if any, α-subtype histone mRNA in large oocytes. These results suggest that these maternal mRNAs accumulate in the pronucleus of the haploid egg after completion of meiotic maturation where they are stored until their release during the breakdown of the pronucleus during prophase.

SummaryThe primary (animal-vegetal) (AV) and secondary (oral-aboral) (OA) axes of sea urchin embryos are established by distinct regulatory pathways. However, because experimental perturbations of AV patterning also invariably disrupt OA patterning and radialize the embryo, these two axes must be mechanistically linked. Here we show that FoxQ2, which is progressively restricted to the animal plate during cleavage stages, provides this linkage. When AV patterning is prevented by blocking the nuclear function of β-catenin, the animal plate where FoxQ2 is expressed expands throughout the future ectoderm, and expression of nodal, which initiates OA polarity, is blocked. Surprisingly, nodal transcription and OA differentiation are rescued simply by inhibiting FoxQ2 translation. Therefore, restriction of FoxQ2 to the animal plate is a crucial element of canonical Wnt signaling that coordinates patterning along the AV axis with the initiation of OA specification.

We present an initial characterization of a database that contains temporal expression profiles of sequences found in 35,282 gene predictions within the sea urchin genome. The relative RNA abundance for each sequence was determined at 5 key stages of development using high-density oligonucleotide microarrays that were hybridized with populations of polyA+ RNA sequence. These stages were two-cell, which represents maternal RNA, early blastula, the time at which major tissue territories are specified, early and late gastrula, during which important morphogenetic events occur, and the pluteus larva, which marks the culmination of pre-feeding embryogenesis. We provide evidence that the microarray reliably reports the temporal profiles for the large majority of predicted genes, as shown by comparison to data for many genes with known expression patterns. The sensitivity of this assay allows detection of mRNAs whose concentration is only several hundred copies/embryo. The temporal expression profiles indicate that 5% of the gene predictions encode mRNAs that are found only in the maternal population while 24% are embryo-specific. Further, we find that the concentration of >80% of different mRNAs is modulated by more than a factor of 3 during development. Along with the annotated sea urchin genome sequence and the whole-genome tiling array (the transcriptome, Samanta, M., Tongprasit, W., Istrrail, S., Cameron, R., Tu, Q., Davidson, E., Stolc, V., in press. A high-resolution transcriptome map of the sea urchin embryo. Science), this database proves a valuable resource for designing experiments to test the function of specific genes during development.

We have determined spatial patterns of expression of individual actin genes in embryos of the sea urchin Strongylocentrotus purpuratus. Radioactively labeled probes specific for each of five cytoplasmic-type (Cy) and the single muscle-type (M) mRNAs were hybridized in situ to sections of fixed embryos. M actin mRNA appears only late in development and is confined to a few cells associated with the coelomic rudiments. The five Cy mRNAs fall into three sets, whose times and sites of expression during development are highly distinctive. Different cell lineages express messages of one or more of these sets, but never all three. Although all Cy actin mRNAs exhibit monophasic accumulation in the RNA of whole embryos during the course of development, such accumulation in many cases results from the summation of both increases and decreases in abundance within individual sets of cells. Within the genomic linkage group CyI-CyIIa-CyIIb, expression of CyI and CyIIb appears to be co-ordinate, and quite distinct from that of CyIIa. CyI and CyIIb are expressed in all lineages at some point in embryogenesis, but confined mainly to oral ectoderm and portions of the gut of the pluteus larva. CyIIa mRNAs are restricted to mesenchyme lineages throughout late gastrula stage, and subsequently accumulate in parts of the gut. The CyIIIa and CyIIIb genes, which form a separate linkage group, are expressed only in aboral ectoderm and its precursors. Furthermore, CyIII messages are the only detectable actin mRNAs in this cell lineage after late blastula stage.

Data on sequence organization in metazoa are reviewed and tabulated. It is shown that the features of sequence organization previously observed in Xenopus DNA are extremely widespread. At least 70% of DNA fragments 2,000–3,000 nucleotides long contain both single copy and repetitive sequence in all the organisms examined except Drosophila.

A sea urchin (Strongylocentrotus purpuratus) messenger RNA encoding a protein (SpEGF2) related to epidermal growth factor (EGF) was identified. The full-length complementary DNA sequence predicts a protein with an unusually simple structure, including four tandem EGF-like repeats and a hydrophobic leader, but lacking a potential transmembrane domain. Sequence similarities suggest that the peptides are homologous to two peptides from a different sea urchin species, which cause a classic developmental defect, exogastrulation, when added to the seawater outside of embryos. The SpEGF2 messenger RNA begins to accumulate at blastula stage, and in pluteus larvae it is distributed in discrete regions of ectoderm that are not congruent with known histological borders. One region corresponds to that expressing the homeodomain-containing protein, SpHbox1. The structure of the SpEGF2 protein and the pattern of accumulation of its messenger RNA suggest that it may have important functions as a secreted factor during development of sea urchin embryos.

RNA transcripts complementary to a genomic histone repeat are found in high concentration in sea urchin egg pronuclei. In situ hybridizations with the recombinant plasmid pCO2 indicate that the nuclear concentration is at least 25 to 50 fold higher than that in the cytoplasm. If nuclear transcripts are predominantly histone mRNAs, they comprise about 12% of the histone mRNA in eggs, or about 0.36 pg. After fertilization these molecules persist through pronuclear fusion but disappear from nuclei by mid 2-cell stage. A similar high nuclear concentration is not observed for polyadenylated mRNAs or for two individual abundant maternal mRNAs. The high steady-state concentration of nuclear histone repeat transcripts suggests that they have an unusually long lifetime in pronuclei of unfertilized sea urchin eggs.

Measurements are presented of the number of transcripts in sea urchin embryos and in adult intestine cells complementary to two cloned genomic single copy sequences. These sequences apparently code for maternal mRNAs stored in the egg. About 1000 and 1400 transcripts are stored per egg, close to the average value for maternal mRNAs. RNAs complementary to the cloned sequences are also found in the polysomal message of early embryos. The number of polysomal transcripts representing each cloned sequence is typical of the “complex” or “rare” class of message in sea urchin embryos. Expression of these putative complex class structural genes is regulated developmentally. Thus, polysomal transcripts of both sequences essentially disappear by the blastula stage. One sequence is again represented at a low level in adult intestine cytoplasmic RNA while the other is absent. However, in all the developmental stages examined, the number of nuclear transcripts of the cloned sequences remains about the same as for any average single copy sequence transcript.

We present an improved procedure for detecting poly A tracts in situ by hybridization of 3H poly U. Glutaraldehyde fixation achieves significantly higher retention of RNA and better morphologic preservation than does Carnoy's. A dramatic increase in signal to noise is obtained by prehybridization treatment of glutaraldehyde-fixed sections with proteinase K and acetic anhydride. Measurement of the increase in poly A concentration after fertilization by solution titration and by in situ hybridization are in excellent agreement indicating that in situ measurements yield accurate relative estimates of local RNA concentrations in sections. Examination of the grain density distribution in section of sea urchin eggs and cleaving embryos reveals no major cytoplasmic localization of poly A+ RNA, although nuclei show much less labelling and micromeres of 16-cell embryos have a small, but significant, reduction in poly A concentration.

The fraction of ribosomes loaded on polysomes is about 95% in logarithmically growing Tetrahymena thermophila, and about 4% in starved cells. Cytoplasmic extracts from cells in these two physiological states were used to develop column chromatographic methods for the purification of polysomes. Bio-Gel A 1.5 m was found to separate total cytoplasmic ribosomes from many soluble proteins, including RNAse, with no detectable change in the polysome size distribution. Polysomes can be separated from monosomes and non-polysomal mRNA by chromatography on Bio-Gel A 15 m without size selection. These methods can easily be adapted to large scale preparations of polysomes, even from cells where a small fraction of the ribosomes is on polysomes. A method is described for reversible precipitation of polysomes and monosomes from dilute solutions at pH 5.3 which greatly facilitates polysome isolation. Hybridization of 3H-labeled polyU to RNA isolated from column fractions has been used to demonstrate that purification of EDTA released polysomal mRNA can be performed using the column chromatography procedures described here. These methods have been employed to demonstrate that most of the cytoplasmic mRNA in log-phase Tetrahymena is loaded onto polysomes while most of the mRNA is starved cells exists in a non-polysomal form.

Arylsulfatases are a group of enzymes that remove sulfate moieties from a diverse set of substrates including glycoproteins, steroids, and cerebrosides. We have isolated recombinant cDNA clones corresponding to an arylsulfatase (SpARS) message that encodes an abundant protein of pluteus larvae of the sea urchin Strongylocentrotus purpuratus. Although vertebrate arylsulfatases have broad tissue distributions, in situ hybridization with a probe for SpARS shows that the sea urchin message accumulates in the embryo only in the single cell type of aboral ectoderm and its precursors. The message is first detectable by RNase protection assays around hatching blastula stage and accumulates through pluteus larva stage. The open reading frame of cDNA clones is 1701 nt long and encodes a deduced protein with a predicted molecular mass of 61 kDa. Analysis of corresponding genomic DNA clones reveals that the pre-mRNA contains six exons. Consistent with the fact that arylsulfatase enzyme activity is extracellular, this polypeptide has a hydrophobic leader sequence and three potential glycosylation sites. Furthermore, hybridization in situ shows that in blastulae arylsulfatase message is preferentially concentrated around nuclei at the basal sides of cells. The S. purpuratus sequence is very similar to that recently reported for the same enzyme from Hemicentrotus pulcherrimus and 30% of the amino acid residues are also identical to those of both human arylsulfatase C (steroid sulfatase) and arylsulfatase A. Sequence relationships among these four mRNAs suggest that, assuming equal rates of evolution, the duplication separating the human genes occurred at about the time of separation of the echinoderm and vertebrate lineages.

We have examined the importance of the extracellular environment on the ability of separated cells of sea urchin embryos (Strongylocentrotus purpuratus) to carry out patterns of mRNA accumulation and decay characteristic of intact embryos. Embryos were dissociated into individual blastomeres at 16-cell stage and maintained in calcium-free sea water so that daughter cells continuously separated. Levels of eleven different mRNAs in these cells were compared to those in control embryos when the latter reached mesenchyme blastula stage, by which time cells in major regions of the intact embryo have assumed distinctive patterns of message accumulation. Abrogation of interactions among cells resulted in marked differences in accumulation and/or turnover of the individual mRNAs, which are expressed with diverse temporal and spatial patterns of prevalence in intact embryos. In general, separated cells are competent to execute initial events of mRNA accumulation and decay that occur uniformly in most or all blastomeres of the intact embryo and are likely to be regulated by maternal molecules. The ability of separated cells to accumulate mRNAs that appear slightly later in development depends upon the presumptive tissue in which a given mRNA is found in the normal embryo. Messages that normally accumulate in cells at the vegetal pole also accumulate in dissociated cells either at nearly normal levels or at increased levels. In one such case, that of actin Cylla, which is normally restricted to mesenchyme cells, in situ hybridization demonstrates that the fraction of dissociated cells expressing this message is 4- to 5-fold higher than in the normal embryo. In contrast, separated cells accumulate significant levels of a message expressed uniformly in the early ectoderm but are unable to execute accumulation and decay of different messages that distinguish oral and aboral ectodermal regions. These data are consistent with the idea that interactions among cells in the intact embryo are important for both positive and negative control of expression of different genes that are early indicators of the specification of cell fate.

Previous studies demonstrated that the pronucleus of the unfertilized sea urchin egg contains a high concentration of transcripts complementary to the early histone repeat unit (D. L. Venezky, L. M. Angerer, and R. C. Angerer (1981). Cell 24, 385-391.) In this paper, in situ hybridization techniques of improved sensitivity are used to show that these nuclear RNAs include authentic histone mRNA but not spacer-complementary sequences. It is estimated that most early-variant histone mRNA contained in the egg is, in fact, restricted to the pronucleus. These mRNAs are released to the cytoplasm at the time of nuclear breakdown of first cleavage and rapidly distribute throughout the cytoplasm.

We have developed techniques for localization of individual mRNA species in sections of sea urchin embryos using the hybridization of radioactively labeled RNA probes to tissue sections in situ. The method yields grain densities proportional to the concentration of target mRNAs and is sufficiently sensitive to allow the localization of most moderately abundant mRNAs. The major features of the method are briefly discussed, and several examples of its use are reviewed. We show that maternal histone mRNA is largely sequestered in pronuclei of unfertilized eggs and released at nuclear membrane breakdown of first cell division. We discuss several events in differentiation of cells of the dorsal ectoderm, including expression of a cell-type specific set of messenger RNAs.

In situ hybridization has emerged as a valuable tool for the identification of individual cells expressing specific genes. Recently, methods have become sufficiently sensitive to detect mRNAs present at the level of only a few molecules per cell. When mRNAs are expressed in only a small fraction of the cells in a mixed population, in situ hybridization may be the most sensitive nucleic acid hybridization technique available. The authors' laboratory has shown that antisense RNA probes offer a unique combination of advantages for detection of individual mRNAs by in situ hybridization. Most importantly, antisense probes provide a large increase in sensitivity due to the absence of competing probe self-reassociation. The high stability of RNA-RNA duplexes allows use of higher post hybridization wash temperatures to achieve a given fidelity of base pairing (stringency), which also reduced backgrounds. RNA transcripts can be synthesized from truncated templates such that they are essentially devoid of vector sequences. This sequence purity maximizes the signal to noise ratio since lower probe concentrations are required to saturate target RNAs.

A decade has passed since the first report of the detection of an mRNA by in situ hybridization (1, 2). In the past five years progress in two areas has greatly increased the utility of this technique. First, the technology of molecular genetics has made available a large number of gene sequences for use as probes for specific cellular and viral mRNAs. Second, work by a number of different investigators has led to the development of in situ hybridization techniques for an increasing variety of biological materials. These efforts have provided characterizations of the technique with respect to important experimental variables, increased the sensitivity of detection, and documented the specificity of the method. In situ hybridization is one of few methods which permit examination of gene expression at the resolution of single cells. It has demonstrated or potential applications ranging from basic problems in molecular genetics and developmental biology to use in diagnostic medicine.

Patients with the acquired immunodeficiency syndrome (AIDS) are subject to a spectrum of central nervous system (CNS) disorders. Recent evidence implicates the human T-cell lymphotropic virus type III (HTLV-III) in the pathogenesis of some of these illnesses, although the cells infected by the virus have yet to be identified. Using in situ hybridization, we examined brain tissue from two patients with AIDS encephalopathy for the presence of HTLV-III RNA. In both cases, viral RNA was detected and concentrated in, though not limited to, the white matter. The CNS cells most frequently infected included macrophages, pleomorphic microglia, and multinucleated giant cells. Less frequently, cells morphologically consistent with astrocytes, oligodendroglia, and rarely neurons were also infected. The findings strengthen the association of HTLV-III with the pathogenesis of AIDS encephalopathy. In situ hybridization can be applied to routinely prepared biopsy tissue in the diagnosis of HTLV-III infection of the CNS. ( JAMA 1986;256:2360-2364)