Casey Luskin recently participated in a debate, on YouTube’s NonSequitur Show, with Daniel Stern Cardinale, an evolutionary biologist at Rutgers University who is also known as “Dr. Dan” of the YouTube channel Creation Myths. It was a cordial, friendly exchange, and I commend Dr. Dan for his willingness to engage in civil and amicable discourse. Both sides handled themselves well, and the dialogue was substantive and illuminating:
ID and Junk DNA
In Luskin’s opening statement, he explained the concept of “junk DNA” and discussed how proponents of intelligent design (ID) have historically predicted that the preponderance of our genome would be determined to be functional, rather than being nonfunctional debris that has accumulated over the eons of evolutionary history. ID does not entail that all of our genome must be functional (since genes or components may be rendered nonfunctional by inactivating mutations), but it does lead us to anticipate that much of our genome will, in fact, be functional. Luskin also reviewed how empirical data have confirmed ID’s predictions concerning “junk DNA” over the last ten or fifteen years — in particular, the evidence of mass transcription that has been uncovered by ENCODE. He pointed to a literature review compiled by Richard Sternberg, Casey Luskin, and myself which documented over 800 peer-reviewed papers showing function for junk DNA. This collection is only the tip of the iceberg, but it shows the radical change in our knowledge over the past couple decades pointing toward functions for junk DNA. Luskin’s main point is that intelligent design predicted this revolution in our understanding of function for non-coding or “junk” DNA, whereas biologists operating from an evolutionary view failed to anticipate this paradigm shift.
Dr. Dan does not deny the evidence that the great majority of our genome is transcribed into RNA or that we are aware of many examples of function for “junk.” Rather, his response to these data is to argue that most of the RNA being transcribed is nonfunctional junk. In his opening statement, the main basis he cited for this presumption is a 2014 paper by Kellis et al. which noted that while “RNA transcripts of some kind can be detected from ∼75% of the genome,” nonetheless “a significant portion of these are of low abundance.” [1] Indeed, “For polyadenylated RNA, where it is possible to estimate abundance levels, 70% of the documented coverage is below approximately one transcript per cell.” [2]
Luskin Responds
After Dr. Dan’s opening statement, Dr. Luskin immediately offered a twofold response. First, he correctly noted that this is an average — and, therefore, a mean of one transcript per cell does not imply that every cell only has one transcript or less. Luskin noted that ENCODE studied 147 cell types, and it’s well known many non-coding DNA elements are only active and functional in certain cell types at certain stages of the life cycle. Thus, if 145 of the cell types studied recorded no transcripts for a particular element, but two of them recorded 50 transcripts each, the average might be less than one transcript per cell and yet the element is highly active in certain cell types.
Second, even in cells where there is a low copy number of RNA transcripts, genetic elements producing those transcripts can nonetheless be functional. Luskin cited a 2022 article in Nature Methods which noted that although “One criticism long leveled at the ncRNA [non-coding RNA] field is that, given their low abundance and low expression, ncRNAs can’t be that important,” nonetheless various scientists have shown that non-coding RNAs “can punch above their weight,” and act in a nonstoichiometric way to amplify effects.” He further cited a paper by Mercer et al. 2012 in Nature Biotechnology which noted that functional long non-coding RNAs (lncRNAs) “were present at an average of ~0.0006 transcripts per cell, indicating expression in only a small subpopulation of the cells sampled” — exactly the kind of pattern Luskin said is compatible with function. They note that even some Hox regulatory gene transcripts may be present at a rate of “an average ~0.13 transcripts per cell” — and these are obviously functional transcripts.
John Mattick and Paulo Amaral
He also cited a 2023 book by John Mattick and Paulo Amaral which explains that lncRNAs can be functional even when transcribed at very low levels [3]:
[L]ow-level and fragmentary signals from lncRNAs in sequencing datasets is mainly a consequence of their highly developmental stage-specific expression, exacerbated by insufficient sequencing depth, especially in complex tissues. The expectation that high expression levels reflect functionality is based on the prevalence of protein-coding RNAs, which are, on average, more highly expressed than regulatory RNAs, although there are exceptions. Indeed, mRNAs encoding regulatory proteins such as transcription factors are usually expressed at lower levels than those encoding structural or metabolic proteins, and have shorter half-lives. [Emphasis added, internal citations omitted.]
They also highlight examples of regulatory RNAs that are expressed only at a low copy number per cell [4]:
Regulatory RNAs would likewise require relatively low average expression levels, i.e., more localized expression, and more dynamic control. Examples, among many others, include functionally validated chromatin-associated lncRNAs detected on average in less than ten copies per cell in populations. Single-molecule RNA FISH revealed that the localized TERT (telomerase reverse transcriptase) pre-mRNA occurs in 9–10 copies per cell and is only spliced during mitosis. Similar low expression levels are also observed for a number of functionally well validated regulatory RNAs, including XIST.
Even one of the papers cited by Kellis et al. — in support of the statement that “70% of the documented coverage is below approximately one transcript per cell” — observes that [5],
…except for IRF4, which was usually expressed at several dozen copies, most [general transcription] factors were detected at <10 copies per cell, and were often not detected at all. We stress that this does not mean that they are not expressed. Given the 10% psmc of the protocol, these observations are consistent with simple technical failure to detect them. It is also possible that there are no mRNA copies in some cells at the moment of harvest, especially if they are infrequently transcribed. Extending these observations to other functional groups, we assessed proteins involved in translation (as a major group of genes with housekeeping functions) (Fig. 3F), splicing regulators (Fig. 3G), and all transcription factors (Fig. 3H). The median number of copies per cell was ∼100 for translation proteins, ∼10 for splicing regulators, and strikingly, only ∼3 for transcription factors. Beyond their biological interest, these large expression differences between functional gene categories mean that quantification is inherently less robust and less informative for some biological functions than it is for others. [Emphasis added.]
Thus, Luskin successfully refuted Dr. Dan’s primary argument in the debate (that low transcription copy number strongly supports a lack of function), and it was notable that Dr. Dan curiously dropped this argument for the remainder of the debate.
Dr. Dan’s Other Argument
For much of the rest of the dialogue, Dr. Dan leaned primarily on his other argument — that is, that the genes encoded by repetitive DNA (especially LINE elements) are too degraded to serve a useful function. During the debate Luskin rebutted this presumption by providing examples of degraded repetitive DNA which can have function. But we have since done another literature review which shows extensive evidence of function for what Dr. Dan called “degraded” LINE sequences. For our rebuttal to this argument, we refer readers to an article I co-authored with Richard Sternberg and Casey Luskin.
Dr. Dan put out his own brief reflections on the debate, which you can find here. There, he highlighted a remark made by Luskin in his concluding statement that scientists have not, as yet, identified specific functions for the majority of the human genome. Dr. Dan apparently thought this was a major concession. But we have never claimed otherwise, and Luskin in fact stated this upfront in his opening statement — fully acknowledging that there is much we don’t know about the genome. Rather, we would contend that ENCODE’s findings that most of our genome is transcribed is, in itself, prima facie evidence of function. Moreover, given the trends of scientific research in recent decades (which have uncovered more and more function for supposed “junk” DNA), this provides grounds for confidence that these trends will likely continue.
A Good Faith Attempt
Again, I commend Dr. Dan for being an amicable dialogue partner and for his good faith attempt to engage substantively with the relevant scientific issues. Judging a “winner” for any debate is usually a complex undertaking, since this will be perceived differently by various observers depending on educational level and background beliefs (among other factors). Indeed, I am biased in that Luskin and I are friends and colleagues. But I thought that he had good counterpoints to all of Dr. Dan’s key arguments in the debate. Though Dr. Dan is correct that we currently know of specific functions for significantly less than half of the genome, this is hardly a strong argument for supposing that the “dark regions” of the genome are non-functional “junk” — particularly given the trends in the scientific literature over the last couple of decades — and the fact that the great majority of our genome is transcribed.
Notes
1. Kellis M, Wold B, Snyder MP, Bernstein BE, Kundaje A, Marinov GK, Ward LD, Birney E, Crawford GE, Dekker J, Dunham I, Elnitski LL, Farnham PJ, Feingold EA, Gerstein M, Giddings MC, Gilbert DM, Gingeras TR, Green ED, Guigo R, Hubbard T, Kent J, Lieb JD, Myers RM, Pazin MJ, Ren B, Stamatoyannopoulos JA, Weng Z, White KP, Hardison RC. Defining functional DNA elements in the human genome. Proc Natl Acad Sci USA. 2014 Apr 29;111(17):6131-8.
2. Ibid.
3. John Mattick and Paulo Amaral, “RNA the Epicenter of Genetic Information,” CRC Press, 2022, p. 157.
4. Ibid.
5. Marinov GK, Williams BA, McCue K, Schroth GP, Gertz J, Myers RM, Wold BJ. From single-cell to cell-pool transcriptomes: stochasticity in gene expression and RNA splicing. Genome Res. 2014 Mar; 24(3):496-510.
This article was originally published, on March 16th, 2024, at Evolution News & Science Today.