Shapiro begins in chapter one with some case studies illustrating the limitations of the Central Dogma: how E. coli chooses which sugars to eat, DNA proofreading, DNA damage and mutagenesis, cell cycle checkpoints, pheromone responses in yeast, and the role of intercellular signals in cell death.
Chapter two, "The Genome as a Read-Write Storage System", argues that "gene" is inconsistent or poorly defined. It then looks at the processes of genome compaction, chromatin formatting, and epigenetic regulation, and their roles in replication and transmission to daughter cells. There follows a survey of the different mechanisms of "natural genetic engineering": DNA exchange with the environment, homologous recombination, non-homologous end-joining, site-specific reciprocal recombination, DNA transposons, long terminal repeat (and non-LTR) retroelements, retrosplicing group II introns, inteins, and diversity-generating retroelements. Examples of these mechanisms in use include the generation of phase and antigen variation, yeast mating types, macronucleate ciliate protozoa, and the mammalian immune system — and there's a survey of mechanisms for their regulation and targeting.
This is presented rapidly and succinctly. Here's an example.
"CRISPR stands for clustered regularly interspaced short palindromic repeats. CRIPSRs were initially identified by bioinformatic analysis that annotated their repetitive structures. It turns out that the active components are actually the spacer elements between the CRISPR repeats. As in piRNA-encoding loci, the CRISPR spacers accumulate fragments of viruses, plasmids, and other genome invaders. Although prokaryotes do not seem to possess the same kinds of epigenetic silencing machinery as eukaryotes, protein sequence motifs and recent experiments indicate that spacer transcripts produce small-silencing RNA molecules (siRNAs) that prevent virus replication. In addition, experiments demonstrate that bacteria sensitive to particular viruses can become resistant by incorporating viral sequence fragments as CRISPR spacers."
And much of this material is presented even more compactly, in tables. "Various Stimuli Documented to Activate Natural Genetic Engineering", for example, lists "natural genetic engineering functions" induced by different "signals or conditions" in different organisms.
This, and the rest of the book, is solidly referenced, with over sixty pages of references and more available online. Shapiro attempts to make Evolution accessible to non-specialists by including references to popular level presentations, mostly from Scientific American, and providing a twenty-five page glossary that includes such basic terms as "diploid" and "peptide bond". There are no illustrations or diagrams, however, and the presentation is rather concise. Readers will need a decent grasp of basic genetics and cell biology to make much sense of Evolution.
Chapter three examines some "higher level" discoveries from molecular genetics and genome sequencing: antibiotic resistance and horizontal gene transfer; proteins as modular combinations of domains; Archaea as the third domain of life; symbiogenesis in the origins of eukaryotes; natural genetic engineering in genomic innovation; the use and reuse of innovations; the epigenetic and regulatory changes at the heart of species differences; and whole genome doubling as a key driver of evolutionary change.
There is some speculative material in this — Margulis' hypothesis that microtubule-based cilia derive from endosymbiont spirochete bacteria, for example, and Donald Williamson's "larval transfer" idea — but it is mostly marked as such. Chapter four, however, is pitched as "going beyond Darwinism" and makes much broader and more radical claims about evolutionary theory. And here I largely part company with Shapiro.
He wants to relegate selection from being the key driving force of evolution. The material he has marshalled, however, weighs against the reification of a single level of "the gene" as the only locus of selection, not against selection more generally. Selection is just as important at other levels: some genomic changes, whether reformattings or insertions or duplications or shufflings or deletions or regulatory reconfigurations, will be viable and successful; others will not.
Shapiro is also unhappy with so much being attributed to "random, accidental genome change". Again, the evidence he marshals seems compelling against the idea that uniformly distributed single nucleotide mutations are the be-all and end-all of evolutionary variation, but not an argument against a central role for stochastic processes more broadly. The kind of genome changes he catalogues may be canalised by all kinds of constraints, but are hardly deterministic.
When it comes to a longer-term palaeontological perspective, Shapiro emphasizes the importance of extinction events and is, not surprisingly, a fan of punctuated equilibrium. But he leans much further towards saltationism than most. It is one thing highlighting genome duplication as a method of speciation, another denying the possibility of gradual speciation outright, with a claim that "selection has never led to formation of a new species". (Perhaps "allopatric speciation" and "character displacement" should be added to the glossary.)
As an alternative Shapiro appeals to systems biology. He doesn't offer any kind of integrated theoretical framework here, however, only a collection of suggestions — about reorganisation of existing functions, duplication and diversification, and "targeted genome restructuring" — which again reflect the importance of different levels of organisation rather than being replacements for variation and selection. For the origin of novelty he suggests that, perhaps, "viruses serve as sources of novelties that can later be adapted by cells", but that would simply relocate the problem.
Rather than leaping straight to extinction events and a complete rewriting of evolutionary biology, the most immediate implications of a broader understanding of genomics seem to me to lie in cell and developmental biology. Here, for a perspective on selection and variation at a range of levels, and on the balance between contingency and constraint and between flexibility and conservation, I recommend John Gerhart and Marc Kirschner's Cells, Embryos, and Evolution. This presents a hierarchical perspective on cell biology and development, exploring the complexities of variation and selection at different levels.
Two minor issues. Firstly, Shapiro draws some analogies with computing which I find rather awkward. "Proteins operate as conditional microprocessors in regulatory circuits. ... the LacI repressor is a microprocessor modulating gene expression in response to the inducer". Such individual cell components are hardly complex enough to be comparable to microprocessors, however, and it would be an unnatural complication to model cell regulatory networks as distributed algorithms where nodes can carry out arbitrary computations.
Secondly, I will indulge myself with a brief rant about the title. There are books called Evolution which are entirely about biogeography and palaeontology, in which nothing molecular features at all, so this is hardly an informative title. And the subtitle is provocative — since the trend so far in 21st century biology has been towards more and faster sequencing, keeping the emphasis on genomes as nucleotide sequences — but again not particularly descriptive. I think a better choice would have been something like The Read-Write Genome: Evolution and Natural Genetic Engineering.
Even those who disagree with most of it may get something out of Evolution: A View from the 21st Century. If nothing else, chapter two makes a nice survey of some of the complexities of genome structure and change which are obscured by a sequence-level perspective. It is unlikely to appeal to readers of popular science, however: they will find the first part of it too dry and read the second part as support for intelligent design.