Quotes of interest — beware single citations and non-citations.

As readers who have been following the Quotes of interest series will know, I have been arguing that from the discovery of repetitive DNA until at least the mid-1980s, the general expectation was that it must somehow be functional for the organism. By 1989 or 1990, we start to see claims that noncoding sequences were “long dismissed” as mere “junk” or “parasites”, and that biologists finally are beginning to recognize that it is interesting. We have heard this line, more or less unaltered, ever since. This statement is usually not backed up with any reference to the literature, it is more of a “general sense” sort of claim. Everybody knows junk DNA was dismissed as irrelevant, so using the standard statement to that effect need not be supported by any actual citations. Either that, or authors who make the statement will cite Ohno (1972), and possibly also Doolittle and Sapienza (1980) and Orgel and Crick (1980), even though these original publications did not in any way dismiss possible functions for some noncoding elements.

This trend of picking one or two references to cite (if any) in support of the claim that all of biology neglected the possible functions of noncoding DNA goes back nearly two decades. I have already provided this quote in a previous post, but I want to use it as an example of how single citations (not to mention non-citations) can be extraordinarily misleading about the past state of the field when it comes to this issue. In 1990, Willard argued that,

Although it has been recognized for over 20 years that the centromeric heterochromatin in chromosomes from virtually all complex eukaryotic organisms consists of various families of satellite DNA, they have only recently been taken seriously as candidates for something other than ‘junk’ DNA or genomic ‘flotsam and jetsam’ (Miklos 1985).

I think it is obvious from the summary of the literature on satellite DNA that the claim that its possible role in chromosomal structure was neglected is incorrect. But what of the citation of Miklos (1985)? Did he represent at least a minority of biologists who long dismissed satellite DNA as nothing more than “junk” or “flotsam and jetsam”?

As it turns out, I have already cited a few of Miklos’s publications from the 1970s and 1980s. Here are some highlights:

Yamamoto and Miklos (1978):

The most important aspect of satellite DNA remains the nature of its functions. Although a large body of data has been gathered concerning its structure, distribution and properties in several different organisms, most of these results have in fact neither supported nor disproved any one of the particular hypotheses of function. The most popular hypothesis on satellite DNA function has been, and still is, that satellite DNA is involved in some aspect of chromosome mechanics such as chromosome pairing.

John and Miklos (1979):

Thus the large amount of satellite DNA in some species, its apparent rigid conservation in sequence and, as we shall see, its effects on the genome when it is altered in amount or position lead us to be unimpressed in general with the argument that most of it constitutes a functionless burden which many eukaryotes must bear. However, for the moment we will retain an open mind and examine the hard data pertaining to function before casting a final judgment at the end of this article.

So, it would appear that Miklos had an interest in the possible functions of satellite DNA, and he certainly took the possibility seriously that they could be important for the chromosome. By 1985, his view had changed to a large extent on this, but this was due to his evaluation of the data, not a long-held assumption about nonfunction.

As he wrote in 1985,

It is this blatant profligacy of repeated sequences within some eukaryotic genomes that fascinates many investigators, and has led to two general and mutually exclusive conclusions concerning their status in the genome. One is that these sequences have functions; the other is that they represent the natural outcomes of genomic turnover processes that involve selfish, ignorant, or junk DNA. The functionalists rationalize the existence of these sequences have been held to function, for example, in determining centromere strength, chromosome pairing, recombination, the three-dimensional architecture of the nucleus, genomic reorganization, and speciation.

On the other hand the proponents of the second class of hypotheses propose that these sequences are inevtiable byproducts of molecular mechanisms involved in DNA replication and recombination, and that such genomic turnover is not supervised by natural selection acting through the phenotype.

These viewpoints accurately reflect other deep divisions in contemporary biology, such as those concerning selection versus neutrality and function versus junk. An overt unease is now apparent in studies on highly repetitive DNAs, since, despite focusing the awesome power of recombinant DNA technology onto the structure of these sequences, the expected biological roles have still proved elusive.

The whole problem of whether there was a function at all for highly repetitive sequences exploded into the harsh glare of criticism when the concept of selfish DNA was propounded. Doolittle and Sapienza (1980), presenting arguments based on the properties of mobile elements, and Orgel and Crick (1980), using similar arguments pertinent to highly repetitive DNAs, argued that sequences having little effect on the phenotype could be spread throughout a genome simply on the basis of their preferential replicative properties.

I have spelled out these classical hypotheses because they can be compared with the themes of the 1970s and 1980s, which are expressed in terms of DNA sequences. Thus satellite or highly repetitive DNAs have been invoked in chromosome pairing, chromosomal rearrangements, speciation, germ line processes, alteration of regulatory pathways via genomic reorganization, and the three-dimensional structure of the interphase nucleus. Highly repetitive DNA sequences may, alternatively, be thought of as examples of selfish DNA or ignorant DNA. Thus highly repetitive DNAs are considered to be subject to natural selection, to be neutral, or to be subject to molecular drive.

This litany of functional explanations led Doolittle and Sapienza (1980) and Orgel and Crick (1980) to formulate their concept of selfishness, in which no phenotypic or evolutionary function was called for.

After this introduction, Miklos (1985) spent 58 pages reviewing the available data on highly repetitive DNA sequences. In the end, it is clear that two major developments influenced Miklos’s views on satellite DNA: 1) despite considerable effort to generate supporting data for proposed functions, these remained lacking in his view, and 2) there was now an alternative, non-adaptive explanation that could account for the existence of much noncoding DNA.

Miklos (1985) further discussed the significance of the new, non-adaptationist, selfish DNA hypothesis near the end of his chapter.

Doolittle and Sapienza (1980), Doolittle (1982), and Orgel and Crick (1980) have drawn attention to what has bogged down our evolutionary concepts of the genome. We have been wedded to thinking of it in functional terms and have considered natural selection as an omnipotent reaper that would rapidly slash any nonproductive DNA from the genome. The existence of highly repetitive sequences has inevtiably led to a search for their function, and this in turn has revolved around a role in chromosome mechanics or chromosome architecture. The data, however, do not support either of these theories. Consequently, evolutionary avenues have been explored. Thus, highly repetitive sequences have been invoked in facilitating chromosome rearrangements, but once again the data base does not favor such excursions. Similarly, a role has been sought for them in speciation processes, but again no unbroken chain of data leads to this as a likely possibility. Doolittle and Sapienza (1980), Doolittle (1982), stressed that most “explanations” are unrealistic. If there are DNA sequences without effects on phenotype, then they can arise, spread, and be maintained in a functional vacuum. The data indicate that satellite sequences per se have no known phenotypic effects.

Miklos was not one to simply dismiss noncoding DNA as “junk” without a thought. If anything, he was among the most ardent advocates of the need for data to test functional hypotheses about noncoding DNA, and he contributed a significant number of these data himself. His view that most noncoding DNA is non-functional for organisms came only after careful evaluation of the data. As he also noted in 1985,

The cloned sequence data together with the biological data are impressive in their implications; they have negated every functional test to which satellite DNA has been put in a mechanistic cellular domain and they have seriously compromised many of the evolutionary concepts pertaining to this class of DNA. In this field there has been an almost overpowering reluctance to lay aside hypotheses that were erected in the absence of relevant data. There has been a desire to avoid the unpalatable, or the unthinkable — that tandem arrays of highly repetitive sequences have no functional characteristics whatsoever but are the ignorant playthings of genomic turnover processes. There is a fear that if not one [adaptive] hypothesis is left standing, then little of substance will remain in this field. However, highly repetitive DNAs have provided a vanguard for the problems that must be encountered in interfacing molecular biology with developmental and evolutionary biology at any level. These sequences have revealed many of the hurdles likely to be encountered in both coding and noncoding parts of the genome. These problems will not go away, nor will our understanding of genomic relevance be helped if the hypotheses that are erected are of little direct predictive value.

I believe that an acceptable state of affairs has been reached for highly repetitive localized arrays at the cellular level. Most hypotheses have been dispensed with, and we are confronted with the stark realism that the obsessive hunt for function has been tangential and illusory.

Revising one’s view in the face of new theoretical explanations and accumulated data (and indeed, producing some of the data oneself) is exactly how science is supposed to work.


Part of the Quotes of interest series.

Doolittle, W.F. and C. Sapienza. 1980. Selfish genes, the phenotype paradigm and genome evolution. Nature 284: 601-603.

John, B. and G.L.G. Miklos. 1979. Functional aspects of satellite DNA and heterochromatin. International Review of Cytology 58: 1-114.

Miklos, G.L.G. 1985. Localized highly repetitive DNA sequences in vertebrate and invertebrate genomes. In Molecular Evolutionary Genetics (ed. R.J. MacIntyre), pp. 241-321. Plenum Press, New York.

Ohno, S. 1972. So much “junk” DNA in our genome. In Evolution of Genetic Systems (ed. H.H. Smith), pp. 366-370. Gordon and Breach, New York.

Orgel, L.E. and F.H.C. Crick. 1980. Selfish DNA: the ultimate parasite. Nature 284: 604-607.

Willard, H.F. 1990. Centromeres of mammalian chromosomes. Trends in Genetics 6: 410-416.

Yamamoto, M. and G.L.G. Miklos. 1978. Genetic studies on heterochromatin in Drosophila melanogaster and their implications for the functions of satellite DNA. Chromosoma 66: 71-98.

One thought on “Quotes of interest — beware single citations and non-citations.

Comments are closed.