Singularities: Landmarks on the Pathway of Life. Christian de Duve. Cambridge University Press, New York, 2005. 258 pp. $48.00 (ISBN 0521841955 cloth).

The Belgian biochemist Christian de Duve won the Nobel Prize in Physiology or Medicine for the discovery of membrane-bound structures called peroxisomes and lysosomes within living cells. These tiny structures may be evidence of ancient endosymbiotic events, when two kinds of cell came together and combined their talents to take evolution in a new direction.

De Duve's interest in the origin of life and its subsequent history began with these discoveries, but it has not ended there. His new book, lucid and superbly organized, surveys the entire history of life, from the first protocells to complex multicellular organisms such as ourselves. Singularities: Landmarks on the Pathway of Life is filled with insights, and should appeal to all readers with a good grounding in biology and biochemistry.

The book is built around a useful classification of the different mechanisms that account for the singularities that have occurred during life's history. The first of these is deterministic necessity—life could have evolved in only one way. Another is a frozen accident—things are the way they are because of chance events that happened at the time of the singularity. And a third is fantastic luck—an event that happens with extreme rarity can catapult life in a new direction.

De Duve's classification scheme includes other singularities in which both chance and natural selection play a role. These consist of various kinds of bottlenecks: selective bottlenecks in which natural selection weeds out less fit lineages, restrictive bottlenecks in which conditions inside the cell or organism determine how evolution occurs, and a category that he calls “pseudo-bottle-necks,”in which what we see is simply the surviving line of many lines that have been lost through chance or attrition. These, he argues, are the workhorses of the evolutionary process and account for the majority of life's diversity. It is usually impossible to distinguish between pseudo-bottlenecks and selective bottlenecks, but there must have been times during the history of life when both have played a role.

He also gives an amused nod to intelligent design. If life on Earth stems from the activities of some cosmic intelligence, then its features should reflect not the cumulative influence of natural selection, but rather the whim of the designer. Thus, if intelligent design is true, we can only hope that the designer was more intelligent than, say, the folks who designed the Edsel.

De Duve examines the roles of biological singularities in the whole sweep of evolution. Much of the book is devoted to a careful examination of the singularities that might have led to the origin of life. Life's origin is of course the elephant in the room—the biggest unanswered question in all of biology. We can trace all of the Earth's life to a common ancestor that probably lived about three billion years ago, but we can catch only glimpses of what life might have been like before that time. What singularities took place?

Were there proteins (or some approximation of proteins) before there was genetic material, and if so, what role did they play? De Duve suggests, as others have done, that a variety of simple proteinlike compounds existed before there were any genes, and that some of these compounds had limited catalytic capabilities. But he does not explore how these molecules might have been organized so that that they could carry out a series of biochemical reactions efficiently. It seems unlikely in the extreme that any one protocell could have had more than a handful of such primitive proteinlike catalytic molecules, and they would have been diluted by a freight of other molecules that weren't much good for anything at all.

Was there natural selection before any genetic material? If so, what was selected? And how could the advantageous characters of a gene-free protocell have been passed down more than a generation or two before being diluted out? De Duve does not confront this latter question directly, but it lies at the heart of scenarios for the origin of life.

Was RNA the first genetic material? And if so, where did it come from? Adenine and the other nitrogenous bases have been synthesized under conditions that might have been locally present on the early Earth, but as de Duve points out, we are totally mystified by how nucleotides might have appeared.

What were the first energy-utilizing pathways? Glycolysis and the Krebs cycle, or electron transport supplied by energy-rich molecules such as hydrogen sulfide and driven by proton gradients across a membrane-bound vesicle? De Duve votes for glycolysis, while Jeff Bada and I have voted for a simple form of electron transport, but it is impossible at the moment to decide which, on the basis of phylogeny, could have come first. Both scenarios have huge problems.

De Duve also examines what phylogenetic analysis can tell us about later singularities, such as the origin of eukaryotic cells and the origins of chloroplasts and mitochondria. His discussion of these matters, about which we have a good deal more evidence than we do about the origin of life, is extremely cogent and interesting. He traces the possible scenarios for eukaryotic cells' acquisition of mitochondria, giving a balanced view that includes possible multiple origins for the hydrogenosomes that are derived from mitochondria, and raises the question of how mitochondria might have survived in eukaryotic cells before there was enough free oxygen for them to carry out their current functions.

De Duve's book is always logical and balanced, and it is scrupulously fair to all the workers in the field. There are, of course, many other things I would have enjoyed having him discuss. For example, while we can only speculate about the singularities of the distant past, it may be possible to examine singularities in present-day experiments on the origin of life. Imagine a chip with thousands of little wells, each occupied by a tiny membrane-bound protocell incorporating different combinations of proteinoids and nucleotide building blocks. Present-day technology, using light-emitting reactions, would allow the experimenter to pick out the protocell that was most efficient at making ATP from ADP, or making a dipeptide from two amino acids. Multiplication of that protocell and transfer of its daughters to a new set of wells might allow us to find out if Darwinian selection can take place even without the aid of genes.

Once or twice de Duve reveals his regrets that more experiments are not being done to investigate such processes. And he makes one giant prediction: He claims that prebiotic chemistry has a large component of deterministic necessity and that life on other worlds will be found to resemble ours closely at the biochemical level. We may soon see how this prediction pans out.