Scientific meetings, like commissioned review articles and symposia commemorating notable anniversaries, allow one to take stock of a field and measure its progress. Between beautifully prepared French meals, the participants at the Retinoids '97 conference, held September 28-October 1, 1997, in Nice, were treated to a broad overview of recent progress on the biology and biochemistry of vitamin A.

It was an auspicious time for retinoid mavens. Not long before the meeting began it was announced that Alfred Sommer was to receive the prestigious Albert Lasker Award. Sommer was honored for his work demonstrating a link between vitamin A deficiency and increased susceptibility to a number of frequently fatal infectious diseases. The World Health Organization estimates that 250 million children are threatened by vitamin A deficiency, and the work of Sommer and others has led to corrective worldwide public health initiatives of diet modification and low-cost vitamin supplements. Additionally, this year marks the tenth anniversary of the identification and cloning of the first members of the family of retinoic acid receptors (RARs), proteins that mediate the action of retinoic acid, the active metabolite of retinol (vitamin A). That achievement went a long way toward making the study of retinoids - vitamin A and its metabolites - a molecular science. The Retinoids '97 conference reflected the continuing interest in the role of RARs in retinoid biology.

The meeting, sponsored by the European Retinoid Research Group, featured 169 participants from 20 countries, who presented their work on a wide variety of problems related to retinoid activity. The keynote lecture was delivered by Pierre Chambon (Université Louis Pasteur, Strasbourg, France), whose group has done pioneering work on the identification and functional analysis of retinoic acid receptors. Retinol is believed to be converted to retinoic acid by two consecutive oxidation steps, from retinol to retinaldehyde to retinoic acid. What then? In 1987, groups led by Chambon and Ronald Evans (Salk Institute for Biological Studies, La Jolla, California) independently identified a nuclear receptor that binds to retinoic acid and, in the liganded state, subsequently binds to a specific sequence (now known as a retinoic acid response element, or RARE) in the regulatory region of target genes, with the result being the activation or repression of transcription. Retinoid-responsive genes are now known to include those encoding growth factors and their receptors, hormones, protein kinases, components of the extracellular matrix, enzymes involved in intermediate metabolism, proto-oncogenes, and Hox genes. This emerging picture, assembled over the last decade, of the broad scope of retinoid-dependent gene activation helps to explain a large body of classical data describing the astonishing array of effects that are observed in animal models and in humans upon vitamin A deficiency or excess. Retinoic acid, with its finger in many different genetic pies, has taken its place alongside vitamin D and thyroid hormone as a potent nuclear-receptor-mediated signaling molecule.

But things are getting more complicated, as evidenced by the number of paradoxes in the literature and in the presentations. To open the meeting, Chambon reported on recent progress in identifying and characterizing RAR cofactors - proteins that appear to be required along with the receptors for retinoid-dependent gene activation. In particular, he presented data on three novel proteins called TIF1, 2, and 3 that bind to retinoic acid receptors and have domains that exhibit similarity to proteins known to be involved in chromatin remodeling. Interestingly, when bound to liganded RARs, TIF1 undergoes autophosphorylation, suggesting that retinoid signaling might also be influenced by phosphorylation through one or more kinase cascades. TIF2 interacts with p300/CBP, a histone acetylase that also has a role in chromatin remodeling. The significance of these interactions is unclear, but the generation of TIF1 and TIF2 knockout mice is underway, and it will be interesting to see if the TIF1 and/or TIF2 null phenotypes resemble those of the retinoic acid receptor null mice.

Several speakers presented new results on retinoid metabolism, particularly on the enzymes that are responsible for generating and clearing retinoic acid. Mary Gamble (of William S. Blaner's lab at Columbia University, New York) discussed the isolation and characterization of a novel 9-cis retinol dehydrogenase, an enzyme involved in the oxidation of 9-cis retinol to 9-cis retinaldehyde. At the other end of the spectrum, Martin Petkovich (Queen's University, Kingston, Canada) reported the cloning and characterization of a retinoic acid-inducible, retinoic acid-metabolizing cytochrome p450 (p450RAI), an enzyme important for the catabolism of retinoic acid. Of particular interest is the expression of this enzyme in the early mouse embryo in the caudal neuropore before neural tube closure. Petkovich speculated that p450RAI might have a protective role in retinoic acid-sensitive tissues prior to neural tube closure.

A major theme in the area of retinoid function was the role of retinoic acid in promoting apoptosis (cell death). The observation that retinoic acid has this effect on a variety of cell lines in culture has been made many times, and we are now getting a clearer picture of how this occurs at the molecular level. Enrico Garattini (Mario Negri Institute for Pharmacological Research, Milan) described the effect of all-trans retinoic acid on acute promyelocytic (APL) blasts. Retinoic acid promotes apoptosis of these cells through the activation of particular members of the caspase family (ICE and Mch-2), proteases involved in promoting apoptosis triggered by a variety of signals. Work along a similar line in Magnus Pfahl's lab (Sidney Kimmel Cancer Center, San Diego) raises the possibility that RAR gamma might be the key receptor in retinoic acid-mediated apoptosis, since an RAR gamma-specific agonist could promote caspase activation and cell death in a number of cancer cell lines. Complicating this picture was Jochen Buck (Cornell University Medical College, New York) who described the activity of anhydroretinol, a so-called "retro-retinoid" (retinoids in which the double bonds are shifted by one carbon atom). Anhydroretinol, too, promotes apoptosis, but it does so by a caspase-independent mechanism, suggesting that there may be many intracellular pathways leading to cell death, each activated by a different subset of retinoids.

John Voorhees (University of Michigan, Ann Arbor) gave an extremely interesting talk that was particularly appropriate for a meeting held in a city with a Mediterranean climate. Retinoic acid can repair photo-aged skin, and Voorhees outlined a pathway through which this repair process might occur. First, retinoic acid can promote collagen expression by transactivation of the collagen gene. Additionally, retinoic acid can inhibit AP-1-driven metalloproteinase expression by suppressing c-jun (a component of AP-1) protein expression. Metalloproteinases such as collagenase promote matrix destruction and are responsible for the appearance of damaged, photo-aged skin. This finding has important implications for the potential use of synthetic retinoids to treat damaged skin, since it is clear that any such retinoid will require both activation and repression functions to be fully effective.

The role of retinoids during embryogenesis was addressed by several participants, beginning with Malcolm Maden (King's College, London), who detailed his recent studies on the vitamin A-deficient quail embryo. In these embryos, of the eight rhombomeres (segments of the hindbrain) that normally form in early-mid-gestation, only the three most rostral rhombomeres actually appear. The tissue that was fated to form the other rhombomeres undergoes cell death, a phenomenon Maden referred to as "positional apoptosis." Maden stressed the striking contrast between the role of retinoic acid in killing many cancer cells and its apparent role as a survival factor in the embryo. Presumably the cellular and developmental context in which the retinoid signal is received determines the outcome, but that is all that can be said for now.

Another study with potentially far-reaching implications was reported by Jacques Berard (University of Sherbrooke, Quebec). He and his colleagues generated a line of transgenic mice expressing an RAR beta antisense mRNA. At 14-18 months of age, these mice exhibited a significant number of lung tumors, indicating that RAR beta is a tumor suppressor. Since RAR beta knockout mice generated in the Chambon laboratory have not been reported to have an increased incidence of lung tumors, Berard suggested that this might be due to upregulation of other RARs in the knockout mice. In contrast, expression of RAR beta in his transgenics was reduced, not eliminated, possibly avoiding the compensatory upregulation of other receptors. Since the human RAR beta gene is located in a region of the human genome that is frequently deleted in lung cancer, these mice may be useful as models to study the role of retinoids in lung tumorigenesis.

Peter Medawar, the Nobel Prize-winning immunologist whose wit and eloquence illuminated so many areas of scientific thought and practice, said about the exponential growth of scientific knowledge:

The ancient Greek Hippocratic school of medicine, according to George Wolf in the July 1996 FASEB Journal, knew at least one big thing in regard to vitamin A: it correctly prescribed raw beef liver, rich in vitamin A, for night blindness and a variety of illnesses now known to be infectious in origin. Now we can swallow vitamin A tablets (thankfully), and we are beginning to discover the basic scientific underpinnings of this treatment, and others like it. Still, there are many paradoxes to be resolved before the field can be said to have reached maturity in the Medawarian sense. Toward that end, the Retinoids '99 conference will be held in Strasbourg, France, where the participants will meet to record the fall of a few more apples.