Biological evolution, part II: Toward an Extended Synthesis24

The Modern Synthesis, forged in the 1930-40s, reigned largely unchallenged until the 1990s. Although challenges to the Modern Synthesis popped up from time to time, most of them had been proven to be either scientifically unsound (e.g., Lysenko’s pseudo-genetics that completely denies Mendelian-Morganian genetics) or at least not firmly established (e.g., the notion that the immune system has typical Lamarckian inheritance, see Steele, Lindely, and Blanden 1999).

Since the 1990s, however, several key new discoveries have completely changed the landscape, and these new discoveries have been confirmed in a wide variety of species, although the exact molecular mechanisms behind some of them (e.g., parental effect, genetic or non-genetic) are not clearly understood yet. Facing these new discoveries as “anomalies” in Kuhn’s (1970) terms, some revision and extension of the Modern Synthesis seems inevitable.

Two far-reaching developments have been behind this new thinking. First, the Modern Synthesis was systematized mostly with physiological data plus limited genetic data from animals and plants at a time with little understanding about the genomics ofboth kingdoms, not to mention other organisms such as microbes (i.e., bacteria, Archaea, and single-celled eukaryotic organisms) and a host of selfish replicons (i.e., viruses, retroviruses, transposons, plasmids, viroids).Today, we have access to a vast amount of genomic data for all these organisms and other life forms and these new data have undermined some of the key assumptions that underpin the Modern Synthesis. Second and directly following the first and Darwin’s “descent with modification (as speciation),” the Modern Synthesis was foremost concerned with speciation too, taking the definition of species to be an easy task. With the vast new genomic data that reveal extensive horizontal genetic transfer and even endosymbiosis across organisms (especially among microbes and selfish replicons), however, it becomes evident that defining species cannot possibly be an easy task, at least for microbes (Doolittle and Zhaxybayeva 2009; cf. Mayr 1997). All these exciting developments call for an extension (at the minimum), if not a fundamental rethinking, of the Modern Synthesis (Wbese and Goldenfeld 2009; Koonin 2009; Koonin and Wolf 2009,2012; Booth, Mariscal, and Doolittle 2016).

Evolution: Darwinian, Neo-Darwinian, Lamarckian, Neo-Lamarckian, or whatever

At the very beginning of this slightly more technical discussion of inheritance in evolution, it is critical to note that the conventional phrasing, “inheritance of acquired characteristics” is utterly imprecise and thus confounding. This is most critically because although “acquired characteristics” within the phrase “inheritance of acquired characteristics” conventionally means only phenotypes, strictly speaking, genetic information or mutations are also “acquired characteristics.”

Back in Lamarck and Darwin’s day, almost everything was lumped under the label of “characteristics,” and there was no distinction of phenotype and genotype; this distinction was not introduced explicitly until 1911 (Johansen 1911). After Mendel, Weismann,Johansen, and Morgan, biologists now explicitly distinguish phenotype from genotype, and this distinction is now a cornerstone of evolutionary biology.²’ This distinction also holds the key to untangling the messy situations caused by the still widely deployed usage of“inheritance of acquired characteristics.”

With the distinction of phenotype from genotype, two things become abundantly clear. What Lamarck in his original “inheritance of acquired characteristics” scheme and Darwin in his pangenesis inheritance scheme had in mind was DIP-WGM. Meanwhile, what the Weissmann Barrier establishes is that, most of the time, DIP-WGM is impossible and only “indirect inheritance of phenotypes via genetic materials” (hereafter, IDIP-VGM) is possible.

Partly because so little was known about genetic material and other fundamental mechanisms of biological evolution during Lamarck and Darwin’s time,²⁶ the possibility of IDIP-WGM was widely accepted, and even Darwin himself was unwilling (or unable) to rule it out. This is the original scheme of Lamarckian inheritance, and Darwin was a believer in it.²⁷ On this front, Darwin differed from Lamarck only at the exact process of, but not on the fundamental possibility of, DIP-WGM. Whereas Lamarck did not have anything concrete to say regarding how direct inheritance of phenotypes can be achieved, Darwin’s pangenesis scheme postulated that every (somatic) cell inoculates a trace of itself into the next generation.

Although Darwin accepted the possibility of DIP-WGM via his scheme of pangenesis, he did differ from Lamarck in a crucial aspect. Lamarck contended that selection proceeds variation or that selection and variation are tightly linked: Environment changes induce or guide adaptive variations in directly. In contrast, Darwin contended that variation presents before the ordering activity of the environment (i.e.,“natural selection”) comes into play and that variation and selection are “de-coupled” (Mayr 1982, 354;Toulmin 1972, 337-338; see also Dunnell 1980,42; Dawkins 1986 [1996]; Kronfeldner 2010,195-198). Although Darwin believed in DIP-WGM just as Lamarck did, Darwin held different positions from Lamarck on other issues, most prominently on the role of natural selection and the relationship between variation and selection. In this sense, Darwin was all by himself as a Darwinian: He was neither strictly “Neo-Darwinian” nor strictly “Lamarckian,” but partially Darwinian and partially Lamarckian, and this is why the modern understanding of biological evolution is called “Neo-Darwinism.”

In recent years, at least four key discoveries in evolutionary biology have made it clear that our rejection of DIP-WGM after Weismann had perhaps been too sweeping and underspecified. There are instances of inheritance that cannot be easily classified as purely Darwinian, even in biological evolution (for a general introduction, see Bonduriansky and Day 2018). Needless to say, all these discoveries hold critical and “far-reaching” implications for understanding social evolution (Miller 2010).

The first discovery is “epigenetic inheritance’’ which literally means inheritance that is outside of the conventional (i.e., Mendelian-Morganian) genetic inheritance (for succinct reviews, see Jaenisch and Bird 2003; Richards 2006; Bird 2007). Abundant evidence now exists that almost all eukaryotic organisms can respond to environmental changes (e.g., diet change and stress) by modifying the genetic materials via (de)methylation of nucleotides and chromatin structures via (de)acetylation of histones without changing the actual sequences of DNA. Indeed, at least some cases of inheritance that were previously identified as conventionally genetic have been shown to be cases of epigenetic inheritance (Danchin et al. 2011,475).

Most critically, at least some of these modifications can be directly transmitted to the next generation through meiosis. Because these modifications change the expression of specific genes, these modifications produce clearly detectable phenotype changes, even though the DNA sequences of these genes remain the same. As a result, these modifications are directly inherited and the phenotype changes entailed by these modifications are indirectly inherited by the next generation.

Yet, although epigenetic inheritance is clearly a form of“soft inheritance” (Mayr 1982; see also Richards 2006; Jablonka and Lamb 2005), epigenetic inheritance cannot be classified as either purely Darwinian or purely Lamarckian. Certainly, because epigenetic modifications still involve modification of genetic materials, epigenetic inheritance is not the original Lamarckian inheritance. Rather, epigenetic inheritance is Neo-Lamarckian (see Table 2.3 for a summary of these notions).

TABLE 2.3 Inheritance of Phenotype: Genetic and Non-Genetic

(Continued)

TABLE 2.3 (Continued)

Phenotypes/traits

Source'. Adapted and modified from Figure 2 and Box 4 of Danchin et al. 2011 and Figure 1.1 of Danchin 2013.

Moreover, the fact that epigenetic inheritance has a very ancient origin (present in all eukaryotic organisms) points to the very possibility that epigenetic inheritance may be an important adaptive trait of eukaryotic organisms. Under many circumstances of stress, an organism that is capable of responding to environmental changes with reversible modifications of genetic materials (and hence also modifications of phenotypes) while maintaining a stable genome holds important advantages over an organism that is incapable of such a response (Danchin et al. 2011,476-477; Halfmann et al. 2012).

Moreover, epigenetic inheritance is not restricted to multicellular organisms. A similar mechanism of epigenetic inheritance, called the CRISPR-Cas system— which stands for “clustered regularly interspaced short palindromic repeats” (CRISPR) and “CRISPR-associated proteins” (Cas), has been found in bacteria and Archaea. Essentially, the CRISPR-Cas system endows bacteria and Archaea with an immune system with acquired defense against invasive genetic elements such as viruses

(or phages) or plasmids.The mechanism of the CRISPR-Cas system is this: Host bacteria or Archaea cells integrate invasive DNA elements into CRISPR loci, and these loci are then transcribed and processed into small interfering RNAs which then guide nucleases for specific cleavage of complementary sequences to silence the invasive DNA elements (for reviews, see Barrangou and Luciano 2014; Sternberg et al. 2016). The CRISPR-Cas system is therefore a form of epigenetic inheritance, with some elements that are not exactly Mendelian and hence possibly “Lamarckian” (Koonin and Wolf2009,2012), although some biologists do not entirely agree (e.g.,Weiss 2015).²⁸

The second discovery is the discovery of prion as the immediate causal pathogen of a family of eventually fatal neural degenerative diseases, including scrapie in sheep, bovine spongiform encephalopathy (BSE, or the “mad-cow disease”) in cows, and Creutzfeldt-Jakob disease (CJD) in humans by Stanley Prusiner and others in the early 1980s (reviewed in Prusiner 1998, 2012). Prion is a protein that takes on a misfolded form (known as PrP^Sl). PrP^Sc can directly cause neural degeneration because PrP^Sc molecules make de novo synthesized (good) prion molecules (known as PrP^c) fold into “bad” prion proteins (i.e., PrP^Sl). As such, a “bad” prion protein can infect its host. At the same time, however, some mutations of the PrP^c gene inevitably lead to the making of PrP^Sc, and this fact makes scrapie, BSE, and CJD inheritable. Obviously, the transmission of these diseases is not (purely) Darwinian, although the inheritance of them is Mendelian-Morganian.

Yet, the transmission and inheritance of scrapie, BSE, and CJD are not purely Lamarckian either. Lamarck’s original formulation insists that the environment induces changes, and these changes are then directly transmitted to the next generation without any genetic materials involved. Although the refolding ofPrP^c into PrP^Sc can be broadly understood to be driven by environmental changes (including the condition of the body), PrP^Sc still has to rely on de novo synthesized PrP^c molecules in order to cause the diseases, and those de novo synthesized PrP^c molecules can only come from transcription and then translation underpinned by the gene that encodes the PrP^c polypeptide. Hence, prion (as a protein) simply does not replicate as a gene does. At the same time, however, some mutations of the PrP^c gene make the diseases as hard as other instances of “hard inheritance.”

Like epigenetic inheritance based on (de)methylation of DNA and (deacetylation of histones,²⁹ prions are also highly conserved: More than two dozen prions have been discovered in yeast (Halfmann et al. 2012; see also Halfmann and Lindquist 2010). Again, this shows that prions may confer an important adaptive advantage to organisms because an organism that can react to environmental changes via modification of prions without compromising its DNA sequences holds important advantages over an organism that is incapable of such a response. Indeed, prions may even have a key role to play in the formation of long-term memory, at least in Aplysia and Drosophila (Si et al. 2010; Majumdar et al. 2012).

Even more intriguing is that epigenetic inheritance and prions can interact with each other, at least in wild-type yeast. Indeed, “prions are a common mechanism for phenotypic inheritance in wild yeast” (Halfmann et al. 2012; see also Halfmann and Lindquist 2010). Again, such an interaction may have an important adaptive advantage because an organism that can react to environmental changes without compromising its DNA sequences holds important advantages over an organism that is incapable of such a response.

The third discovery is the so-called “parental effect” upon progenies, which can be both genetic and non-genetic. Parental effects are “effects that parents have on the phenotypes of their offspring that are unrelated to the offspring’s overall genotype” (Mousseau 1998). Briefly, “parental genetic effects occur when the expression of parental genes in one of the parents becomes an environmental component affecting the development of the offspring” (Danchin et al. 2011,477). In contrast, parental non-genetic effects (PNGEs) operate mainly via transmission of cytoplasmic or somatic factors such as prions, RNAs, hormones, antibodies, and nutrients (Bonduriansky and Day 2009,109-110), and PNGEs can affect other genetic pathways (such as sexual selection) directly or non-genetic pathways (such as propagule size), which in turn would impact genetic pathways such as survival and mating. Indeed, Arai et al. (2009) found that “juvenile enrichment” can rescue a genetic defect in long-term potentiation and memory formation (for a succinct discussion on PNGEs’ far-reaching implications for extended inheritance, see Danchin et al. 2011,477-479; for more detailed discussion, see Hager, Cheverud, and Wolf 2008; Badyaev 2008;Badyaev and Uller 2009).

Finally, there is “niche construction” or “ecological inheritance” (Odling-Smee 1988). Niche construction means that organisms can modify the environment in which they live via their biochemical and physical activities (e.g., worms burrow into the soil, leaving tunnels), and these physical and chemical modifications of the environment can then pass down to the next generation^) and come back to shape organisms’ selection and fitness (e.g., Lewontin 1983; Laland, Odling-Smee, and Feldman 1996, 2000; Odling-Smee, Laland, and Feldman 2003; Odling-Smee 2010; Danchin et al. 2011; Kendal,Tehrani, and Odling-Smee 2011).³⁰ Obviously, the presence of “niche construction” implies that an organism’s environment is not independent from an organism’s life cycle. Rather, there is feedback from an organism’s life cycle to the environment. By any measure, human beings have been the most potent and versatile in all animal species when it comes to changing their environment (Kendal et al. 2011; Odling-Smee and Laland 2011).The coming of settled agriculture, the Industrial Revolution, mega-cities, the internet revolution, and global warming are just a few examples.

Altogether, these new discoveries imply that the transmission of instruction or information (with gene being only one form) and phenotype in biological evolution is far more complex than the gene-centric and externalist Modern Synthesis had anticipated (for a summary, see Table 2.4). Put differently, the gene-centric and externalist view of evolution, especially variation and inheritance via DNA replication, is thus incomplete, to say the least (Bonduriansky and Day 2009,2018; Pigliucci and Muller 2010,12-14; Danchin 2013).

Thus, although an “Extended (Evolution) Synthesis” is not in place yet, several different pathways have been suggested (e.g., Carroll R. L. 2000; Carroll S. B. 2008; Helantera and Uller 2010; Pigliucci 2007; Pigliucci & Müller 2010; Danchin et al. 2011; Kendal et al. 2011; Danchin 2013), and the central tenets of the Modern Synthesis are valid for many (if not most) circumstances in biological evolution, one

TABLE 2.4 An Information-centric versus a Gene-centric Vocabulary of Evolution

thing is certain:The notion that the gene-centric and externalist “Modern Synthesis” has solved all the mysteries of biological evolution is no longer tenable (Kendal et al. 2011). As such, we may need to forge a more inclusive theory of variation and inheritance and hence the whole evolution process (Blute 2017).

Indeed, some of the core components for a new “Extended Synthesis” are now firmly in place. Increasingly, biologists are no longer bound by the gene-centric and externalist view of variation and inheritance and are moving toward a more information-centric and interactionist view of variation and inheritance, with genes being only one carrier of information (see especially Wagner and Danchin 2010; Danchin 2013). This information-centric and interactionist “Extended Synthesis” may demand a whole new set of vocabulary that is more accommodating than the vocabulary of the gene-centric and externalist “Modern Synthesis” (see Table 2.5 for an illustration). As becomes clear in the chapters that follow, an “Extended Synthesis” about biological evolution poses extensive and profound implications for understanding social evolution.

Finally, our increasing recognition of the complexity of biological evolution, especially its variation, selection, and inheritance, has made the practice of labeling evolution as a whole as Darwinian, Lamarckian, and Spencerian increasingly unhelpful, if not counterproductive. At the very least, the coming of the Extended Synthesis makes the dogmatic stand of“(Generalized) Darwinism” for understanding social evolution untenable (e.g., Hodgson and Knudsen 2010).This fact points to an inevitable conclusion: We have to drop the practice of labeling evolution in whole as Darwinian or Lamarckian, even for rhetorical purposes.³¹ As becomes clear in Chapter 3 and Chapter 4 later, it makes virtually zero sense to debate whether social evolution is Darwinian, Lamarckian, or whatever (Lewens 2015; Tang 2017; cf. Hodgson and Knudsen 2010).