CHAPTER 1
Introduction
Except for Cassandra herself, prophets of doom have almost always turned out to be wrong, especially where scientific developments are concerned, and our remarks are not meant to discourage the great enthusiasm with which many molecular biologists are abandoning K12 for BALB C or some other mammal, such as a nematode. (F. Jacob and J. Monod. 1970, p. 3)
In the initial letter of invitation addressed to me two years ago, the Chairman of the Carter-Wallace Lectureship Committee explained that the purpose of these lectures was "to call the attention of students and faculty in biochemistry, biology, and chemistry to exciting developments in important areas of the life sciences." I assumed that the development to which I owed the invitation was hybridization of somatic cells, and suggested this title for my lectures; and I take the renewed invitation to lecture on this subject to mean that this theme had not come to be regarded meanwhile as less exciting, and hence, that the discouragement and pessimism which afflict nowadays so many biologists do not prevail at Princeton.
This pessimistic attitude was, as some of you may remember, signaled by Sol Spiegelman in the late fifties, when he stated that "the outlook is depressingly bright for the quick resolution of many interesting problems" and, as you know, reached its apogee in Gunther Stent's most articulate book "The Coming of the Golden Age: a View of the End of Progress" in which he claims that most of the essential principles of biology have been discovered and most of the essential problems solved; and that the few remaining ones are either not worth our bother or may not be solvable.
It appears to me true that, for the first time in the history of science, we have recently come to a point wherefrom one can see that science can eventually reach an end, but I do think that Stent's pessimism is premature, and I fully agree with the more constructive views recently expressed in two remarkable articles published in Nature: one entitled "Molecular Biology in the Year 2000" by Francis Crick, the other "Genes and Hereditary Characteristics" by Alfred D. Hershey. I highly recommend these two articles to those who are afraid that we shall soon run out of important problems, and I hope that they will feel somewhat reassured also by what I shall say in these lectures about the status of some fundamental biological questions.
Now, a few words about the plan of my lectures.
In the 10 years since its discovery in 1960, hybridization of somatic cells has grown from a biological curiosity into a method of analysis which is now so widely used in investigations of the genetic basis of a variety of biological phenomena, that an exhaustive review of the literature in a few lectures is no longer possible. I therefore think that it will be more rewarding if I limit my objectives: I shall first give you an historical account of the development of the method of somatic hybridization and of our knowledge of the properties of hybrid cells; and shall then speak in a somewhat more detailed way about three areas of biological research, presently investigated by means of somatic hybridization, which appear to me of particular importance or are of particular interest to me personally: formal genetics, cell differentiation, and cancer (Note 1).
CHAPTER 2
History of Somatic Hybridization
I suspect that if we were honest we would have to admit that if any one of us had never been, our science would not have been quite the same; but it would be awfully hard to see the difference. (Sir Macfarlane Burnet. 1967, p. 4)
I chose to start with an historical account of somatic cell hybridization for two reasons.
First, in my opinion, an honest historical account by an eyewitness inevitably shows the respective roles, in any important scientific development, of deliberate, logical design and of "lucky accidents" and, hence, especially in periods of pessimism about the future which many biologists are going through right now, inspires confidence and hope that the unexpected will, again and again, bring unforeseeable solutions to what, on logical grounds, appear today to be insoluble problems.
The second, and much less important reason, is that a history of somatic hybridization has been written recently by Henry Harris, but his version is different from mine, and I may as well take this opportunity to present my own.
In presenting it to you, I shall assume that you are all acquainted, at least superficially, with the principles and techniques of cell culture in vitro on which hybridization of somatic cells relies, worked out chiefly by H. Eagle and T. T. Puck in this country. (The relevant references can be found in the excellent reviews by Green and Todaro and by Krooth et al., and in the book of Morgan Harris; a very brief account is given in Ref. 72. The definitions of some technical terms will be given in the Appendix.)
Since some of you may not be familiar with hybridization of somatic cells, I would like to begin by saying that the occurrence of this phenomenon was discovered in 1960 by a research team working in Paris under the leadership of Georges Barski, and not by me as has been often erroneously stated in reviews of the subject. I hasten to add that one person has not shared in this error: having made, in 1965, an important contribution to the field, Henry Harris, whose style sometimes makes one believe that he may be short on modesty, seems to ascribe the discovery of cell hybridization neither to Barski, nor to me, but to himself (Note 2).
The discovery of hybridization between cells of permanent lines
How was somatic hybridization discovered?
For several years, Barski had been studying two permanent, heteroploid lines of mouse cells growing in vitro and isolated many years earlier by Catherine Sanford and co-workers at N.I.H. (Note 3). These two cell lines were derived from the same initial culture, i.e. were very closely related, but, in the course of years, had become very different in the degree of their neoplasticity. Having observed that the cells of the two lines differed both in their morphology and in their karyotype, Barski decided to look for evidence of Pneumococcus-like transformation by growing the two cell types together: what he apparently was looking for were morphological changes of cells whose identity could be established by their unchanged karyotype. What he discovered was quite different, however. It was the appearance, after three months of mixed culture, of a new cell type which was characterized quantitatively by a total number of chromosomes nearly equal to the sum of the modal chromosome numbers of each of the parental lines, and qualitatively by the presence of essentially all of the marker chromosomes of the two parents, i.e. of the chromosomes which differentiated one cell type from the other. These were, clearly, hybrid cells: they contained in a single nucleus the combined genomes of the two parents.
For a long time, karyological analysis was the only means of identifying hybrid cells, and, as you will see, the correlation of phenotypic traits with the karyotype is one of the most important aspects of somatic cell genetics; therefore, I think that it is worth showing you three microphotographs which illustrate Barski's first experiments (Note 4).
Figure 2.1 shows the karyotypes of the parental cells of this "cross" and of the resulting hybrids. In 2.1A is shown a metaphase of a cell of (the highly malignant) line N-I, with 54 acrocentric chromosomes, including an extra long one. Figure 2.1B shows a metaphase of a cell of the other "parent" line, N-2 (of low malignancy): it contains 61 chromosomes, 12 of which are biarmed (or metacentric). Figure 2.1C shows a metaphase of a hybrid cell with a total number of 113 chromosomes; among them you will recognize the marker chromosomes of this cross: the extra long acrocentric chromosome of N-I and the 12 biarmed chromosomes of N-2.
Barksi's first publications of these results were received with great scepticism. Because of their apparent rarity, hybrid metaphases, observed alongside metaphases of the parental cells, were regarded by most karyologists as artifacts resulting from the superposition of two parental metaphases in the karyological preparations. However, it appeared to me that, if true, Barski's findings were of sufficient potential importance to deserve verification. This I set out to do with my co-workers in early 1960, using two close relatives of the cell lines used by Barski, and within a few months we reproduced Barski's results.
Generality of the phenomenon
Having convinced myself of the reality of the phenomenon, having furthermore isolated pure hybrid populations from this cross (this was a relatively easy task since the hybrids had a selective advantage over the parental cells), and having observed some of the properties of the hybrid cells — I set out with great optimism to prove that somatic hybridization can be forged into a tool of genetic analysis of somatic cells.
What had to be shown to that end, as I saw it then, could be stated in three points. It had to be proved that:
1. It is possible to obtain hybrids between any two cells differing in some or many of their genetic properties.
2. Such hybrid cells can be isolated or preferably selected and maintained in pure culture so as to make the study of their phenotypes possible.
3. The formation of hybrid cells is followed by some process — be it only mitotic abnormalities resulting in the accidental loss of chromosomes — which leads to the formation of cells with different constellations of genes, that is (using the term very loosely), in gene or chromosome segregation.
I must admit that my optimism was based on the rather shaky ground of our first observations of the few somatic hybrids obtained. By now, however, all of these conditions have, to a great extent, been fulfilled, and I will now describe the different steps whereby somatic hybridization reached its present status.
Concerning the first point (i.e. the possibility of obtaining hybrids between genetically different cells), it became clear within the next couple of years that the "spontaneous" formation of hybrid cells, containing in a single nucleus the quasi-complete genomes of both parents, occurs, as a rare event, in mixed cultures of practically any two types of karyologically distinguishable heteroploid cells of the so-called "permanent lines," and that these hybrids can be isolated, either owing to their selective advantage or by cloning. This was first shown by crosses between cells of several pairs of permanent mouse lines of different origins performed in my laboratory (review in Ref. 61). Thus, by 1963, we had growing as pure cultures a number of karyologically identified hybrids between genetically different cells: I say "genetically different" because the cell lines involved had been derived from inbred strains of mice differing in histocompatibility genes and, surely, many other genes. This, in turn, permitted the first studies of the expression, in the hybrids, of the genes by which the parental cells differed. It was thus shown, to begin with, that as expected, these hybrids exhibit the genetic traits of both parents. For example, hybrids produced by the cross of cells having two different H-2 histocompatibility antigen complexes, carry, on their surfaces, the antigens characteristic of both parents: there is co-dominance of the genes specifying these antigens. Similarly, by crossing two mouse cell lines, derived from mice in which the enzyme β-glucuronidase differs in heat stability, hybrids were obtained which synthesized both "allelic" forms of β-glucuronidase. These were the first proofs that both parental genomes are active in somatic hybrids, and numerous similar observations have been made in subsequent years.
Crosses involving normal diploid cells
What I said about the quasi-general occurrence of hybridization between cells of permanent cell lines should not make you think that every cross we attempted resulted in success. We did have a certain number of failures. Possibly they were due to the rarity of hybrids between some cell types or, more probably, to the lack of selective advantage of certain hybrids over the parental cells. Indeed, we learned later by pure accident that many young hybrids, for some unknown reason, are incapable, to begin with, of overgrowing the parental cells at 37°, while they do so at 29° Several new hybrids were isolated in this way, including one of a new type which was of particular importance. It was a first hybrid between cells of a permanent line of mouse cells (2472-6, derived from N-1, already familiar to you) and diploid cells from a newborn mouse carrying a small chromosome translocation (T-6). The karyotypes of the parents and of the hybrid are shown in Figure 2.2. The metaphase shown in A is that of a cell of line 2472-6. It comprises 50 chromosomes among which you will recognize the extra long acrocentric chromosome of N-I. You will notice that it also contains now a long metacentric chromosome which provides another marker in the cross with the diploid mouse cells. The karyotype of the latter, shown in B, comprises the normal complement of 40 chromosomes, all clearly acrocentric except for the two very characteristic "translocated" (T-6) chromosomes. The marker chromosomes of both parents are easily recognized in the metaphase of the hybrid cell, shown in C. The total chromosome number here is 89, which is very nearly the expected one.
This hybrid was of particular importance because it showed, for the first time, that diploid cells can be hybridized also, at least when their partners are cells of a permanent, neoplastic line. Moreover, similar crosses performed not with freshly explanted T-6 diploid cells, but with senescent cells of the same origin, hardly dividing any more, gave rise to similar hybrids. In both cases, they were shown to inherit from the permanent parent the lack of contact inhibition (or, in other words, the ability of multilayered growth, characteristic of permanent lines) as well as the capacity of rapid and continuous growth. The question of the malignancy of such hybrids will be briefly discussed in my last lecture.
A first selective system for the isolation of somatic hybrids
Returning to the selection of hybrids at low temperature: in view of several successes of this method, I at one time had high hopes that we had "put our finger" on a universal system for the selection of hybrids. These hopes did not materialize, however. But the mention of this fact brings me to the next important development — the demonstration by John Littlefield in 1964 that by using two cell lines, carrying different selective markers, one can, as in microbial genetics, establish a selective system wherein the hybrid cells will grow and the parental cells will not. Littlefield's system is, in fact, an adaptation to cell hybridization of a selective system designed by Szybalski and co-workers for other purposes.
Littlefield selected two sublines of the heteroploid mouse line known as the L-line: one resistant to 8-azaguanine, the other to 5-bromodeoxyuridine (BUdR). These two drug resistances are correlated, respectively, with deficiencies for the enzymes hypoxanthineguanine phosphorybosyl transferase (HGPRT) and thymidine kinase (TK) which are required for phosphorylation (and therefore for incorporation) of the base analogs. The system of hybrid selection is based on the fact that mammalian cells have open to them two pathways of synthesis of nucleotides (Figure 2.3): the de novo pathway whereby nucleotides are synthesized from sugars and amino acids, and the "scavenger" pathway which utilizes the preformed nucleosides, hypoxanthine and thymidine. The de novo pathway can be blocked by aminopterin. The operation of the scavenger pathway depends on the simultaneous presence of TK and HGPRT. Therefore, the drug resistant (enzyme deficient) parental cells are unable to grow in a medium (HAT) containing hypoxanthine, aminopterin, and thymidine. On the contrary, the hybrids which contain the genes of both parents, and therefore produce both TK and HGPRT, grow unhampered in HAT (Figure 2.3B).
