In this article we will discuss about:- 1. Introduction to the Action of Genes in Differentiation 2. Signals to Genes in Differentiation 3. Recognition of Genes by Signal Molecules 4. Control of Genes at the Transcription Stage 5. Post-Transcriptional Control of Gene Expression 6. Further Possibilities 7. Relation of Differentiation to Mitosis and a Few Others.
Contents:
- Introduction to the Action of Genes in Differentiation
- Signals to Genes in Differentiation
- Recognition of Genes by Signal Molecules
- Control of Genes at the Transcription Stage
- Post-Transcriptional Control of Gene Expression
- Further Possibilities of Genes
- Relation of Differentiation to Mitosis
- Numbers of Genes in Differentiation
- The Time Factor in Progressive Differentiation
1. Introduction to the Action of Genes in Differentiation:
We now have to consider a very serious contradiction between the properties of cleavage nuclei and the fates of cells differentiating in various directions. It has been shown that during cleavage and also in subsequent stages the nuclei of all the cells of the embryo retains the ability to support the development of a whole embryo with all its various differentiations.
ADVERTISEMENTS:
When nuclei of cells are taken from swimming frog larvae or even from adult animals and transplanted into enucleated eggs, the eggs can develop into perfectly normal frogs which do not lack any kind of tissues or cells. This is a clear indication that the cells of animals which have advanced far in development still contain in their nuclei the full set of information necessary for any differentiation occurring normally in that particular kind of animal.
The same conclusion may be reached by considering the processes of regeneration and asexual reproduction. It will be shown that in the repair of damage to the body of adult animals and in the production of new individuals from somatic cells, already differentiated cells may participate in the production of tissues of a different type.
If we accept—as it appears that we must—that the whole genotype with a complete set of information covering all possible differentiations of a certain kind of animal is contained in the nucleus of every cell, or at least in most cells, even when the adult stage is reached, then how is it that in any given cell only part of its potentialities is revealed, while others do not manifest themselves? In other words, how is differentiation of cells possible?
One possible solution of this contradiction is that while the nucleus of every normal cell contains the full complement of genes, not all of them are in an active state. In terms of the mechanism of gene action, this means that not all genes are producing messenger RNA in any given tissue at any given time. The transition of a gene from an inactive to an active state may well be called “activation.” Gene regulation by this mechanism is referred to as regulation at the transcriptional stage.
ADVERTISEMENTS:
The RNA transcript of a DNA sequence in the genome, produced by the action of RNA polymerase, is, however, not the mRNA which directs protein synthesis in the cytoplasm of the cell. The original transcript must be processed, certain parts of the transcript eliminated, and the RNA molecule must be “capped” and adenylated, before the finally processed mRNA passes into the cytoplasm.
This processing may well be selective, so that only the RNA copies of some genes reach the cytoplasm, while the other genes, although active in the sense that they are being copied by the RNA polymerase, do not affect the protein synthesis in the cell, because their RNA copies are not processed to become effective mRNA.
It has been suggested that all genes (the whole length of the DNA strands) are transcribed into RNA at all times, but that the RNA produced by some genes is rapidly destroyed and does not leave the nucleus, while the RNA’s modeled on other genes are suitably processed, and only these RNA’s are passed into the cytoplasm. If this interpretation were correct, the regulation of gene action would be performed not at the transcription stage but subsequently between transcription and translation.
The end result, as far as cytoplasmic differentiation is concerned, would be exactly the same – inactive genes would be those whose complementary RNA is synthesized but is destroyed within the nucleus; the active genes would be those whose RNA is processed and passed into the cytoplasm. Gene regulation of this latter type would be referred to as regulation at the post-transcriptional stage.
ADVERTISEMENTS:
Whatever the mechanism of gene action regulation (transcriptional, or post-transcriptional), it is most likely that the initial factor determining the kinds of genes to be put in action reaches the nucleus from the cytoplasm, or from outside the cell via the cytoplasm.
This follows from the fact that in early development the differences in the differentiation of cells depend on the kinds of cytoplasm that are present in the different parts of the egg and not on the nuclei. Furthermore, the behavior of cells in many tissues is influenced by extraneous chemical substances, such as hormones, which of course can reach the nucleus only by first entering the cytoplasm.
2. Signals to Genes in Differentiation:
What is the ultimate cause of different parts of the embryo developing along divergent lines? One fact as mentioned above is that cytoplasmic differences in the egg determine which path differentiation will take in the various parts of the embryo.
As the differentiation of cells in various parts of the embryo involves activation of genes responsible for the manufacture of corresponding kinds of messenger RNA, it follows that the kinds of cytoplasm present in the cells determine which genes are to be activated.
Nuclear transplantations present an excellent proof that the cytoplasm is responsible for the kind of activity performed by the chromosomes and thus by the genes. To what was stated on transplantation of nuclei into an egg, it may be added that nuclei can also be transplanted into amphibian ovarian oocytes, which present them with an essentially different environment.
An oocyte nucleus does not replicate and therefore does not synthesize DNA. On the other hand, it synthesizes large quantities of RNA, partly ribosomal RNA and partly messenger RNA. Nuclei transplanted into oocytes, like those transplanted into an egg, conform in their behavior to their surrounding cytoplasm.
This is shown to best advantage by transplanting a cleavage stage nucleus into an oocyte. Cleavage nuclei synthesize much DNA and little or no RNA. Transplanted into oocytes, such nuclei stop DNA production and start producing RNA instead. Furthermore, at the time of maturation divisions, transplanted nuclei enter into mitosis synchronously with the host nucleus of the oocyte.
Control of the nucleus as a whole by the cytoplasm makes it necessary to assume that some substances pass from the cytoplasm into the nucleus which causes the chromosomes and the genes to change their activity, repress some genes and activate others. In fact, passage of proteins from the cytoplasm into the nucleus has actually been shown to take place.
ADVERTISEMENTS:
Instead of transplanting nuclei into another cell, cytoplasm of a different nature may be added to a cell to test the influence of cytoplasm on the nucleus. If cells of two different kinds are cultivated in vitro together, some of them fuse, producing composite binucleated cells.
Fusion may occur of cells differentiated in different ways, and also of cells of different species. Hybrid cells are produced in the latter case. Mouse liver (hepatoma) cells have been caused to fuse with human leukocytes. Liver cells synthesize blood serum albumin; leukocytes do not do so.
The hybrid mouse liver + human leukocyte cells continued producing mouse blood serum albumin, but also produced some human blood serum albumin (which could be detected by immunological methods). The result shows that the cytoplasm of mouse liver cells activated the genes in human chromosomes to produce mRNA for blood albumin.
The messages regulating the behavior of nuclear genes, however, do not all originate in the cytoplasm of the cell containing the nucleus in question? Differentiation of cells is to a very large extent dependent on influences reaching the cells from outside. Of primary importance in this respect are the phenomena of embryonic induction.
It has been proved that embryonic induction is mediated by the transmission of a substance (a protein or a nucleoprotein) from the inductor to the reacting tissues, as for instance in the induction of the neural plate by the roof of the archenteron in vertebrates. It is an open question, whether the inducing substance emitted by the archenteron reaches the nuclear DNA as such, or whether it only starts a chain reaction in the recipient cells, the end link of which is the signal received by the genes.
There is no doubt, however, that gene behavior is changed in the process of induction; this is evident because induction is suppressed by the action of actinomycin D, thus new mRNA must be synthesized as part of the cell’s reaction. It is also evident that the signal generated by the inducing substance must have entered the nuclei of the reacting cells via their cytoplasm.
In later stages of development and in the adult organism the behavior of the cells, including their genes, may be under the influence of hormones or hormone-like substances, originating sometimes in distant parts of the body. While facts of this kind transgress into the field of adult physiology, one or two cases must be mentioned here.
Erythropoiesis is a continuous process in healthy animals, but it can also be stimulated to a higher level by a hormone-like substance, erythropoietin, which is produced in the kidneys in response to insufficiency of oxygen supply. Treatment of hemopoietic tissue in vivo or in vitro with erythropoietin stimulates the transformation of stem cells into erythrocytes.
The initiation of hemoglobin mRNA production is prevented by actinomycin D, which proves that DNA-dependent mRNA synthesis is part of the cell’s reaction to the external stimulus. Again, the message could reach the nuclear genes only by first entering the cytoplasm of the reacting cells. The reaction of various tissues to stimulation by sex hormones is known to be sensitive to actinomycin D and thus must involve activation of the genes and DNA-dependent synthesis of mRNA.
The general conclusion from the preceding facts is that the genes become active as a result of signals of some kind which reach them from the cytoplasm of the cells. These signals differ depending on the condition of the cells and depending on the environment—the presence of various substances in the cells’ surroundings.
3. Recognition of Genes by Signal Molecules:
In terms of molecular processes, bringing a gene into an active state means allowing a molecule of an enzyme RNA polymerase to attach itself to a DNA molecule at a specific point, from where it starts moving along the length of a DNA molecule, placing the RNA nucleotides in their correct positions and binding them together.
From what has been said about the formation of RNA precursors, it appears that the point of attachment of the RNA polymerase molecule is not within the section of the DNA coding for the meaningful (functioning) RNA sequence, but within a region specially adapted for this purpose—the “spacer section,” the “initiator sequence,” or the sector coding for the later discarded part of the mRNA precursor molecule. Once attached and functioning, the RNA polymerase molecule works its way along the whole length of the sequence, until it is stopped by a “terminator sequence” or its equivalent.
What is required of the signal coming in from the cytoplasm into the nucleus is (1) to recognize a certain sequence on the DNA molecule which precedes the meaningful sequence to be transcribed as RNA, and (2) to enable the RNA polymerase molecule to be attached to the DNA and to start the synthesis of RNA.
It is doubtful that substances in the environment of cells, such as inducing substances, hormones, etc., possess the properties necessary for the recognition of specific sequences on the chromosomes. Rather, it is likely that the signal molecules entering the cells from without become associated with some sort of molecule within the cell which leads the signal to the correct spot on the DNA. J. Bonner (1971) suggested that such a molecule plays the part of a “seeing eye dog” (the analogy is to guide dogs of blind people).
The “seeing eye dog” could be an RNA molecule, a view to which Bonner was inclined, or it could be a purely protein molecule, like the lac repressor. Obviously, the recognition of the “receptor gene” or the “initiator sequence” is the crucial point in understanding the mechanism of gene regulation.
As the specificity of any particular section of the DNA lies in the order in which the four deoxyribonucleotides are arranged along its length, the most plausible way of recognizing this order is by means of another nucleic acid molecule (perhaps an RNA molecule) with a complementary sequence to that of the DNA.
This is what happens in DNA-RNA hybridization. It has been repeatedly proved, however, that a protein molecule can recognize a specific DNA sequence. In the regulation of gene action in bacteria a protein molecule, the “lac repressor,” finds a specific locus on the chromosome and, by attaching itself to it, represses a particular gene.
Studies on the structure and mode of action of actinomycin D, often used to suppress DNA-dependent RNA synthesis, furnish further proof that proteins are able to recognize a specific DNA sequence. Molecules of actinomycin D consist of a “phenoxazone” ring (a triple ring of carbon, nitrogen, and oxygen atoms) to which a loop of five amino acids is attached on each side. The loops are identical, but rotated in respect of each other by 180°.
The exact shape of the molecule and the way it attaches itself to the DNA helix were determined by the x-ray diffraction method. It was found that the phenoxazone ring becomes inserted in between two adjacent deoxynucleotide pairs, and each of the pentapeptide (amino acid) loops establishes firm hydrogen bonds with a guanine molecule above and below the phenoxazone ring.
Because of the shape of the actinomycin D molecule, it can only “fit in” between two guanosine-cytosine groups, provided that these groups are in the “trans” position, that is, that the guanosines (and cytosines) are in opposite DNA chains –
guanosine-cytosine
cytosine-guanosine
It is then this arrangement of four bases that is recognized by the actinomycin D molecule. The attachment of the actinomycin D to the DNA double helix distorts its shape, and makes it impossible for the DNA to serve as template for RNA synthesis. It is noteworthy that the actinomycin molecule attaches itself to an intact double helix, without the two strands of DNA becoming separated.
Evidently, enough of the steric structure of the nucleotides is accessible in between the “backbones” (sugar-phosphate chains) of the DNA double helix to make it possible for the guanosine-cytosine nucleotides to be recognized. If a small molecule, with partly protein properties (in its pentapeptide loops), can recognize a specific sequence of two nucleotide pairs there is no reason to doubt that a larger protein molecule could recognize a long nucleotide sequence by the same mechanism—that is, by conformation of the arrangement of atoms in the molecules.
Certain enzymes known as endonucleases split DNA molecules at specific points indicated by a particular sequence of nucleotides. The enzyme can thus recognize a specific nucleotide sequence. Lastly, the “splicing enzymes” which eliminate the “introns,” during the processing of the original RNA transcript for mRNA in the nucleus, cut out the introns in such a way that they always begin with the nucleotides GU and end with the nucleotides AG (“Chambon’s rule”). This is a clear indication that the enzyme can recognize a particular nucleotide sequence.
4. Control of Genes at the Transcription Stage:
The most convincing experiment showing gene control at the transcription stage is that done on the ovalbumin gene, which is responsible for the synthesis of vast amounts of ovalbumin in the oviduct of birds under the influence of the female sex hormone, progesterone.
When progesterone reaches the cells of the oviduct (via the blood stream) the ovalbumin genes become active in producing very large amounts of ovalbumin mRNA—up to 10,000 molecules of ovalbumin mRNA per cell in a matter of hours. Subsequently the protein—ovalbumin—is synthesized and secreted into the lumen of the oviduct, to form the main part of the egg white in the eggs, which under normal conditions would be passing through the oviduct at the same time.
In experiments in which the molecules of the hormone were labeled by radioactive hydrogen, it was established that on entering the oviduct cells the hormone molecules are linked with molecules of a special “receptor” protein. The receptor protein has been isolated and purified, and found to be a “dimer,” that is, consisting of two subunits bound together.
Each subunit is roughly cigar-shaped, and has a molecular weight of over 10,000 (by comparison, the molecular weight of the progesterone molecule is only about 300). Two hormone molecules become bound to the two units of the receptor.
Similar protein receptor molecules exist in other tissues which are responsive to steroid hormones—the group of hormones to which progesterone belongs, as well as other sex hormones. The receptors are specific for each hormone, and tissue cells not possessing receptors for a specific hormone do not react to that hormone.
When not in the presence of the hormone, receptor molecules are located in the cytoplasm of the cells. It has been estimated that in the tissue of the rat uterus, which is responsive to estrogen, the receptor molecules for estrogen are present in quantities of about 10,000 per cell.
When hormone becomes available to the cell, and becomes associated with the receptors, the receptor molecule-hormone complex passes from the cytoplasm into the nucleus, and there the complex becomes attached to the chromosomal DNA.
Further stages of the investigation have been carried out in vitro, combining chromatin isolated from nuclei, and in part treated in various ways, with receptor-hormone complexes. It was found that receptor-hormone complexes attach to chromatin from oviduct cells much more actively than they attach to chromatin from other tissues.
Thus the specificity of hormone action is doubly insured- firstly, because not all cells possess the receptors for progesterone, and secondly because the receptors activated by the hormone attach more easily to chromatin of the oviduct cells. Thus, the chromatin of the oviduct cells is somehow different from chromatin from cells in other tissues.
Tests have also been carried out to determine what this difference depends on. When the nuclear chromatin is stripped of all attached proteins, the remaining pure DNA loses its tissue specificity; the receptor-hormone complexes attach in the same degree to DNA’s derived from any tissue other than oviduct tissue.
Thus, the tissue specificity appears to depend on the proteins attached to the DNA. The proteins associated with DNA in the chromosomes are of two main kinds – the histones; and the non-histone-proteins, including the acidic proteins. The histones are very similar in different tissues, and even in different organisms, and thus can hardly be responsible for the tissue specificity of chromatin.
This has been confirmed in respect to the specificity of oviduct cells-stripping the oviduct cell DNA of its histones did not prevent the binding to the DNA of the receptor-hormone complexes. Neither did the replacement of the removed histones by histones extracted from other tissues prevent the DNA-receptor-hormone association.
The non-histone proteins, on the other hand, are much more variable from cell to cell. In oviduct cells there are three groups of non-histone proteins (not more specifically defined), associated with the nuclear DNA. When two of these groups were removed from the DNA, the binding capacity of the chromatin for the receptor-hormone complex was not affected.
When the third group, designated as A3, was removed from the oviduct chromatin, there was a sharp decline in the ability of the remaining chromatin to bind the receptor-hormone complexes. The ability for such binding was restored, however, when A3 proteins were recombined with the DNA.
The last experiment in this series consisted in building “mixed” chromatins, combining the DNA from one tissue and the A3 protein from some other tissue. Oviduct chromatin with non-oviduct A3 proteins lost its ability to bind receptor-hormone complexes, whereas non-oviduct chromatin acquired this ability when its A3 protein was replaced by A3 protein from oviduct cells.
The starting point of this whole investigation was that binding of the protein receptor molecule, with the hormone molecules attached, to the chromosomal DNA causes massive production of ovalbumin messenger RNA (and subsequent secretion of ovalbumin).
It is now evident that for this reaction to take place two prerequisites are needed- firstly, that in the cytoplasm of the oviduct cells there are present the receptor molecules, specific for the hormone progesterone; and, secondly, that the chromatin of the reacting cells includes as a component a specific complex of non-histone proteins.
When both these prerequisites are fulfilled, the ovalbumin gene, which is present in the nuclei of all cells, but quiescent, may become active in the sense that it starts to be transcribed into ovalbumin mRNA. The exact location of the ovalbumin gene on the chromosome is not known, and therefore it is not possible to determine whether the receptor-hormone complex is attached at the location of the gene.
However, this seems highly probable, with the reservation that the active part of the gene (the part coding for the final mRNA) may be preceded by a “leader” sequence of some kind, and that the attachment could take place to the leader sequence.
If the ovalbumin gene could be isolated and cloned, and then tested in a vital system—a frog oocyte—it could be discovered whether the addition of receptor-progesteron complexes are able to accelerate the manufacture of ovalbumin mRNA. A positive result would prove the direct action of the receptor-hormone complex on the gene.
5. Post-Transcriptional Control of Gene Expression:
The opposite view, namely that the intra-nuclear heterogeneous RNA represents a much wider range of information than reaches the cytoplasm in the form of messenger RNA, also has important experimental support.
In one experiment the similarity between active cytoplasmic mRNA and intra-nuclear heterogeneous RNA was tested first in blastula cells from the sea urchin Strorxgylocentrotus purpuratus and then compared with RNA from more advanced stages of development—namely, from the gastrula—and from adult tissues.
Firstly, polyribosomal messenger RNA was isolated from blastula cells and purified. Because the mRNA was isolated from polyribosomes, it is evident that it was being used for transcription into polypeptides and, eventually, proteins. The chromosomal DNA of the blastula stage was broken up into fragments, the two strands of each fragment were separated, and single copy fragments were isolated by allowing multiple copy fragments to hybridize, and then removing the double-stranded fragments of DNA. The DNA was previously heavily labeled with radioactive hydrogen (3H) to allow easy detection and measurement.
Next the cytoplasmic polyribosomal mRNA was allowed to hybridize with the single-copy fragments of the chromosomal DNA, and the resulting double-stranded molecules were separated from the remaining bits of DNA, which did not find counterparts in the cytoplasmic mRNA. The fragments of chromosomal DNA which reacted with the cytoplasmic mRNA could now be assumed to be the sections of the genome, from which the cytoplasmic mRNA had been transcribed.
After separation from the attached mRNA, these fragments could be used as probes, or “tracers,” for testing other samples of RNA. As the fragments of DNA isolated in this way were single copy fragments, they could be regarded as “structural” genes, or parts of the same, coding for mRNA, and eventually for proteins. They could be designated as mDNA.
The sum total of different sequences of chromosomal DNA in the genome of the sea urchin Strongylocentrotus pupuratus has been estimated as being of the order of 610,000,000 nucleotide pairs. By comparison the sum total of different sequences in the polyribosomal mRNA in the blastula cells of the same animal is of the order of 27,000,000 nucleotides.
After protracted hybridization 2.1 per cent of the total of the single-copy DNA sequences reacted with the polyribosomal mRNA, which means that, of all the single-copy sequences of the sea urchin genome, 2.1 per cent are being used in the synthesis of mRNA actually entering the cytoplasm and associating with the ribosomes in the blastula stage.
When the “tracer” mDNA, prepared in the way described, was tested against an excess of polyribosomal mRNA of the blastula, it was found to hybridize with the latter to a degree of 78 per cent. The remaining, non-hybridizing 22 per cent are presumed to be the result of contamination with random DNA sequences that were not completely eliminated during the purification process.
The “tracer” mDNA was then tested against cytoplasmic mRNA prepared from an adult tissue – the intestine of the same species of sea urchin. The proportion of hybridization was found to be much lower, namely 12 per cent. This means that a very large proportion of the gene sequences coding for mRNA in the blastula stage do not produce mRNA in the adult intestine. From other lines of investigation it is concluded that the quantity of different sequences of mRNA in the intestine in terms of nucleotide pairs is equal to 18 per cent of the sequences in the blastula stage.
Thus, in addition to the mRNA’s in the intestine which are identical to those in the blastula, there must be a certain proportion of mRNA’s that are different from those in the blastula and which, presumably, account for the specific differentiation of the intestinal cells. We will note that of the genes giving rise to mRNA in the blastula, only 15 per cent (12 per cent – 78 per cent) give rise to mRNA in the adult intestine.
Next the “tracer” DNA was tested, not against cytoplasmic mRNA, but against the nuclear heterogeneous RNA from different sources. This was derived from gastrula cells, from cells of the adult intestine, and from coelomocytes of the adult animal. The results were quite different from those in which cytoplasmic mRNA was used.
These results are in obvious contradiction to the concept that only those sections of the DNA are transcribed in the nucleus, which are used for the manufacture of the mRNA entering the cytoplasm. We have seen before that parts of the original transcripts from the chromosomal DNA are eliminated in the manufacture of the mRNA (the spacer sequences).
The present data go even further, as they indicate that some sequences of DNA which code for the mRNA at one stage (in the blastula), are transcribed at other stages (in the gastrula and in the adult tissues), although the corresponding mRNA’s do not appear in the cytoplasm of these cells in detectable quantities.
Although this would go beyond what is actually proved, it may be concluded that all or most of the genes are transcribed in the nucleus, giving rise to the heterogeneous nuclear RNA, but only some sequences, different in different types of cells, pass into the cytoplasm as mRNA. The control of gene expression must then lie not at the stage of transcription (all genes being transcribed), but at a later stage, when the heterogeneous RNA is processed in the nucleus, to produce mRNA in part and otherwise to be degraded.
This conclusion is, however, not absolutely binding. Firstly, it may be pointed out that most of the single-copy sequences in the genome, even if transcribed, produce very few RNA copies in the heterogeneous nuclear RNA. These have been calculated as 0.5 – 1 copy per cell in the gastrula, and as little as 1 copy per 10 cells in adult tissues.
For lack of information as to what kind of proteins are coded for by these sequences, it may be questioned whether the small quantities of RNA have any physiological significance, in particular in the differentiation of cells. It may be recalled that in hormone-stimulated oviduct cells the number of mRNA molecules for ovalbumin rises to 10,000 per cell.
The second consideration which has been put forward is that there may be two types of transcription in the nucleus; that while all or most of the genes are being transcribed to produce heterogeneous nuclear RNA, the transcripts destined to become mRNA in a given cell at a specific stage are, right from the start, different in some way from the rest of the nuclear RNA, and that this difference is imposed on the transcripts by some special controlling factor. Thus, the meaningful control would still be at the transcription stage. This explanation of course has to be supported by some further evidence before being accepted.
6. Further Possibilities of Genes:
Transcriptional and post-transcriptional control of gene action does not necessarily exclude one another. It has been suggested, for instance, that while some genes, the products of which are overwhelmingly prevalent in some differentiated cells (such as the genes for ovalbumin or for globin), are activated to a high level at the transcription stage, other genes are all transcribed at the same stable rate, but their RNA transcripts are treated in different ways at the post- transcription stage—as nuclear RNA— with the result that different proportions of individual structural gene transcripts reach the mRNA condition.
It is conceivable, on the other hand, that all genes of a genome are transcribed at all times, but that the rate of transcription is very different in different genes. The genes that are “needed” in a particular tissue, are transcribed at a high rate, and produce massive amounts of mRNA of a specific kind, while others are transcribed at a much slower rate, sometimes perhaps so slowly that the small amounts of mRNA produced cannot be readily detected.
There are numerous reports in the literature that certain kinds of mRNA are produced in cells and tissues in which the protein for which the mRNA codes is “out of place” in view of the function of that cell or tissue. Globin mRNA in particular has been reported as being produced in tissues other than the hemopoietic tissues in which it has a functional significance.
Very substantial quantities of globin mRNA has been found in frog oocytes, and in cells of a malignant tissue known as “Friend cells”. In both these cases globin mRNA was in the cytoplasm, or was found to be adenylated, and thus functional. Globin mRNA also appears to be present in very small quantities in other tissues.
It has been claimed that initially all genes are transcribed, but perhaps at such a low speed that only one molecule of mRNA for each gene is present in 10 to 100 cells. Such small quantities would not be detectable by the usual methods. On the other hand, it is doubtful that the products synthesized on mRNA present in such low quantities could play any role in the functioning of the cells in question.
To become functionally significant the mRNA’s must be produced in much larger quantities, and if so, the change of the gene from the condition in which it is transcribed only once in 10-100 cells, to a condition in which it is transcribed thousands of times in each cell would be rightfully recognized as a process of activation. In this way we are back to the concept of transcriptional control of the function of the genes.
It will be useful to refer here to an experiment in which the transcription of a specific gene was tested in a living system. The genes used were the ribosomal 18S and 28S genes, cloned in a bacterial plasmid. The plasmid DNA included almost the whole 18S and 28S sequences, with their spacers.
It was purified and injected into the nuclei of Xenopus oocytes. The amount of ribosomal gene DNA was about 500 times greater than that of the ribosomal genes in the oocyte itself. After 8-18 hours the nuclear contents of the injected oocytes was spread on grids and examined under the high-power electron microscope.
The injected plasmids were recognizable, owing to their circular structure. The injected DNA was seen to have become reconstituted into chromatin, with nucleosomes in the non-transcribed parts. There was no indication of transcription in the plasmid’s own part of the circle, but in about 10 per cent of the plasmids the ribosomal genes were being actively transcribed.
This was evident from the presence of molecules of RNA polymerase, which are visible in the electron micrographs, and of the threads of RNA spreading out from each polymerase molecule. The whole is exactly similar to what is seen on the loops of the lamp-brush chromosomes in the oocytes, indicating that the “reading” of the ribosomal genes injected into the oocytes was proceeding normally.
It is, remarkable that although some of the injected genes were actively transcribed, others (90 per cent!) were quite inactive. This may be connected with the overabundance of the ribosomal genes after the injection (500-fold compared with the normal amount). It was postulated that there was a shortage of some factor needed to make all the genes active.
This could not be a shortage of RNA polymerase molecules, as one would expect that these would then be sparsely supplied to all the injected genes, instead of being concentrated in the full amount on some genes and avoiding others altogether. The conclusion could be that some other factor, which is in limited supply, is needed to turn on the ribosomal gene, and without which the polymerase molecules do not gain access to it at all.
The experimental system (the frog oocyte) would permit the execution of further tests to identify the unknown factor. If the ribosomal genes could be made to work when injected into the oocyte nucleus, then other cloned genes could also be tested in the same system, and their controlling factors analyzed. It is possible now to pass on to some more speculative concepts concerning the ways by which the action of the genes can be regulated.
The signals activating the genes need not always come directly from outside the nucleus. Britten and Davidson (1969) pointed out that in the process of differentiation not one gene at a time becomes activated, thus causing the synthesis of only one protein, but rather whole batteries of genes come into action, thus accounting for a whole array of enzymes which are characteristic of a particular tissue. Thus the function of the liver in vertebrates involves the production of at least 147 different enzymes, not even counting the enzymes present in each cell, such as those taking part in oxidation, cell growth, and replication.
A single external factor, such as an inducing substance in early embryogenesis or a hormone in later development, must be able to activate more than just one gene. To explain this, Britten and Davidson postulated the existence of a system of regulatory genes. They suggested that the initial signal goes to a “sensor gene,” which through a system of “regulator genes,” attached to it in sequence in the chromosome, produces molecules of “activator RNA.”
These molecules have no other function than to seek out the initiator sequences (or “receptor genes,” in Britten and Davidson’s terminology) of the “producer genes.” The latter are then transcribed into mRNA. A particular producer gene could be called into action starting from several sensor genes, and this would account for the same proteins being synthesized in different types of tissues.
One of the arguments which Britten and Davidson have adduced in support of their theory is that in the cells of multicellular organisms there is far more DNA than is necessary to serve as templates for the mRNA actually used in the synthesis of cytoplasmic proteins.
The rest of the DNA could serve as part of a vast and complicated regulatory system which controls the activity of the mRNA-producing genes. The intra-nuclear fast turnover RNA could in part be made up of the “activator RNA” carrying messages from the sensor system of genes to the producer genes.
More recently this model has been modified by its authors. According to the new model, only a few genes which produce overwhelmingly large amounts of mRNA are actually stimulated to a higher rate of transcription by specific activators. All the rest of the structural genes are supposed to be transcribed at a steady basic rate, but the resulting RNA transcripts are destroyed intranuclearly to different degrees, varying from 100 per cent to 0. The “sensor” genes retain their position and function in the new scheme, and serve for the stimulation of transcription of the “regulator” genes.
The RNA’s transcribed from the regulator genes, however, do not attach themselves to the initiator sequences of the structural genes, so that these genes become activated; instead, the regulator RNA molecules form duplexes with complementary sequences in the “leader” or “trailer” parts of transcripts of the structural genes. The proposers of this new scheme believe that the formation of such duplexes would stabilize the structural gene transcripts, and also serve as signals for the further transformation of the transcripts into mRNA.
The proportions of the transcripts of any particular structural gene which would be stabilized by the formation of duplexes would depend on the numbers of corresponding bits of regulator RNA available at a given time, and this would be determined by the kind of sensor gene activated, and by the number and kind of regulator genes which are under the sensor gene’s control. A sensor gene is supposed to be able to activate any number of similar regulator gene sequences—an idea supported by the existence of multiple copies of some DNA sequences in eukaryotic genomes.
7. Relation of Differentiation to Mitosis:
In a double helix DNA molecule the nitrogenous bases, the arrangement of which provides the specificity of particular parts of the chromosome (genes), are joined to one another pairwise in the interior of the molecule. Many biologists believe that the recognition of the nucleotide sequences would be greatly facilitated if the two strands of DNA were to separate, thus exposing the inner surfaces of the genes.
Such a separation occurs normally in the S phase of the mitotic cycle, when the two strands of the double helix in the chromosome replicate before mitosis. It is therefore very suggestive that the start of differentiation seems to be dependent on a previous mitotic division of the cell.
Actinomycin D prevents the start of hemoglobin synthesis in erythroblasts treated with erythropoietin. Hemoglobin synthesis can, however, also be stopped at an earlier stage by supplying substances which upset the DNA replication, such as fluorodeoxyuridine. It seems therefore that the DNA replication, which normally occurs in the mitotic reproduction of the stem cells, is somehow necessary for the subsequent DNA-dependent mRNA synthesis.
Accordingly, it has also been observed that when erythropoietic tissues are treated by erythropoietin, the first reaction is rapid DNA synthesis which is detectable within the first hour after treatment, whereas hemoglobin synthesis gets underway after two hours and attains its maximum even later.
Very similar observations have been made on mouse pancreatic tissue differentiating in vitro. The production of pancreatic enzymes is prevented if the tissue is treated with actinomycin D in the early stages of differentiation, but once differentiation has got under way, actinomycin D has no effect, showing that mRNA has already been produced in the cells, and protein synthesis can proceed unimpeded.
However, if the pancreatic tissue, at the time when the cells are dividing, is treated with fluorodeoxyuridine, which prevents DNA replication and thus stops cell divisions that normally precede differentiation, the tissue will be unable to differentiate and start its secretory activity.
In numerous other cases differentiation seems to follow a series of mitotic divisions of cells. We may refer to the differentiation of neurons in the central nervous system. There may be the same kind of link between the period of cleavage and the subsequent period of gastrulation, during which extensive gene activation takes place. It is also suggestive that a series of lymphoid cell divisions is somehow involved in the mechanism of production of antibodies in the immunization process.
Lastly, it is possible to quote an observation on a type of cancer, a liver hepatoma. Cells of this tissue grown in vitro are able to synthesize an enzyme, tyrosine aminotransferase, in response to a treatment with corticosteroids. The reaction involves the production of a messenger RNA for the enzyme. The important point, however, is that the induction of tyrosine aminotransferase is possible only during the latter half of the G1 phase and during the S phase of the tumor cells.
This observation, in conjunction with all the facts, makes it plausible that the “seeing eye dog” with the signal molecule attached makes use of the separation of the two strands of the double helix DNA in the chromosomes, replicating in preparation for the next mitosis, to slip in, recognize the gene, and become attached to it.
Now it is believed that of the two strands of DNA of which the double helix is composed; only one strand is used in transcription. Which of the two strands is recognized by the incoming signal? It has been suggested that the signal molecule recognizes and becomes attached to the “passive” strand of the helix.
By blocking one strand, the signal molecule (or the complex containing the signal molecule) would then leave the other strand permanently or semi-permanently open. This strand would then be free to serve as a template for transcription into ribonucleic acid. Thus, the gene would become active. The majority of researchers, however, visualize the attachment of the signal complex to the active strand of the DNA.
8. Numbers of Genes in Differentiation:
One further point has to be discussed in connection with gene regulation in development. In the course of differentiation of certain tissues, some proteins are manufactured in very large quantities in proportion to the overall size of the cell. This is true, for instance, in muscle differentiation, in which up to 10 per cent of the mass of the cell consists of one protein, myosin.
Other cases are hemoglobin in erythrocytes and silk fibroin in the silk gland cells of moths. If the genomes of all cells of one animal are the same, how is it that one pair of genes, constituting only an infinitesimally small proportion of the chromosomal DNA, manages to give rise to a sufficient number of mRNA molecules to account for the synthesis of such prodigious amounts of a particular protein?
We have seen that at least in one case, the oocytes, which have to manufacture very large quantities of ribosomal RNA, this is achieved by means of an amplification of the rRNA-producing genes. Is it possible that in cells synthesizing very large quantities of a few proteins the corresponding genes become amplified, so that not just two, but large numbers of the same kind of gene can be transcribed in order to furnish the necessary quantities of mRNA?
Such a mechanism is feasible in principle, in view of the possibility of gene amplification in the oocytes. The evidence is, however, against the amplification of producer genes. The messenger RNA for silk fibroin has been hybridized with DNA from silk glands on the one hand and with DNA from the rest of the body of the caterpillar on the other.
It was found that the levels of saturation were the same in both cases. This result shows that the number of producer genes for silk fibroin in the silk glands is no greater than the number of these genes in parts of the body which do not produce silk. The results of transplantation of nuclei of differentiated cells into eggs point in the same direction.
If larval gut cells had some of their genes amplified in a way appropriate for gut differentiation, the amplification would have shown itself in some way after implantation into the egg; this was not the case. It appears, therefore, that the cell has some other mechanism for increasing production of proteins starting from only two copies of a gene present in the diploid set of chromosomes.
9. The Time Factor in Progressive Differentiation:
Once the mechanism of gene activation is set in motion, it is not difficult to imagine that when the messenger RNA reaches the cytoplasm it would cause new proteins to be synthesized there. These proteins either could be instrumental themselves in activating further genes, or could, by enzymatic action, produce substances which could act on the genes or their products, thus causing yet new kinds of messenger RNA to be released into the cytoplasm.
In this way more and more genes would become active as development proceeded, and of course the kinds of genes put into an active state would depend upon what direction the differentiation had already taken in the previous stages. More variety in the results of this nucleocytoplasmic interaction would be introduced by the influences of already distinct parts of the embryo on one another, by the activity of “organizers of higher degrees”, by tissue interactions, by the actions of hormones, etc.
The state of a cell is obviously not defined solely by the condition of the genes at any given time (whether a particular set of genes is active or repressed). Rather the previous activities of the genes, as well as the conditions of the environment to which a cell has been exposed, contribute to the state in which the cell is found. These previously existing conditions may in certain cases be decisive and may override subsequent events. One very important contributory factor to this state of affairs is that mRNA produced by the genes and passed into the cytoplasm may sometimes fail to be immediately involved in protein synthesis but begins to be active some time later.
Messenger RNA was first discovered in bacteriophages and was characterized as “fast turnover RNA.” The mRNA of animals is by no means as short-lived and may persist in the cytoplasm for days and possibly even for months. We have seen that messenger RNA produced in the oocytes remains inactive (“masked”) until fertilization occurs, when the oocyte mRNA starts directing protein synthesis throughout the period of cleavage.
There are indications that also in later development mRNA may be produced a considerable time before it becomes active in the synthesis of proteins, which give the visible expression to the differentiation of the respective embryonic parts.
This may be the explanation of determination, which as we have seen often precedes visible differentiation. For instance, in the development of the amphibian embryo the neural plate becomes determined by the middle of gastrulation, when it is under-laid by the chordomesoderm, but the neural plate does not appear until several hours later, in the early neurula stage.
The use of actinomycin to suppress transcription of DNA into RNA and of puromycin to suppress the translation of mRNA into protein structure helps to distinguish between the two processes of determination and differentiation. A particularly clear experiment of this kind has been performed on chick embryos. Treatment of chick embryos with puromycin at the primitive streak stage causes a retardation of development of all systems at the beginning of organogenesis.
The nervous system is thinner than usual, the brain vesicles are reduced in size, and the neural tube is seldom closed. The heart is reduced in size, and often there is no hemoglobin production in the blood islands. Treatment of the embryo in the same primitive streak stage with actinomycin D produces very different results. The neural system is reduced to an extreme degree and in some cases is completely suppressed. The somites are also suppressed, while the heart develops very well and sometimes is the only organ to be differentiated.
Blood islands produce hemoglobin as normal. The fairly obvious interpretation of this experiment is that at the primitive streak stage the mRNA necessary for heart and blood island development has already been produced, but the mRNA necessary for nervous system and somite development has not yet formed, and its production is eliminated by the actinomycin.
The differentiation of the heart, however, takes place later roughly simultaneously with the differentiation of the nervous system and somites, and as differentiation involves protein synthesis, all organs are retarded by puromycin treatment in about the same degree.