How the Vertebrate Brain Regulates Behavior

by Donald Pfaff

  Secondly, the estrogen-binding neurons in ventromedial hypothalamus (VMH) were useful in providing one anchoring point (together with sensory inputs and motor outputs) for working out the lordosis behavior circuit, the first for any vertebrate behavior.

  Third, since the nuclear hormone receptors we study are ligand-activated transcription factors we could use applied molecular biological techniques to demonstrate effects of estrogens on gene expression in behaviorally relevant neurons, especially in VMH (Chapter 3). That work now extends to work in the lab concerning estrogenic effects on histone modifications, epigenetic changes useful for understanding the hormone / behavior relationship.

  Fourth, we used molecular pharmacological manipulations of the VMH and other neurons to prove the gene / behavior causal relations recounted in Chapter 4.

  Taken together, these four advances proved for the first time exactly how specific chemicals acting in specific parts of the vertebrate brain determine a complete behavioral response.

  Reliability

  Of course, I am pleased that a discovery made 50 years ago not only has held up but has expanded into the physiological and molecular explanations of how a vertebrate behavior is regulated. At the molecular level, various discoveries were superimposed on each other to constitute a solid body of understanding. For example, the molecular roles of progesterone receptor (PR) gene expression and its consequences were made clear at the hormone binding level by the estrogenic induction of PR messenger RNA (mRNA) and the subsequent PR gene promoter analysis, a phenomenon that is strongly sexually differentiated (Pfaff 1999, figure 9.1, p. 125). Further, as transcription factors PRs induce new mRNAs, and PR knockout abolishes lordosis behavior. DNA-binding phenomena at the PR promoter are stronger in females than males. The neuroanatomy is right as well: PR is strongly bound in the VMH, and induction of such binding is correlated with lordosis behavior. There, too, estrogens induce PR in females but not males. At these multiple levels of evidence, the molecular biology of estrogen-induced PR gene expression is beyond doubt.

  The neuropeptide enkephalin features the same type of convincing overlay of data (Pfaff 1999, figure 9.1, p. 129); in the VMH, estrogens induce enkephalin mRNA expression through clear promoter binding phenomena. And the VMH result is neuroanatomically specific. Enkephalin’s δ-opioid receptor is induced as well. DNA binding results are stronger in females than males, and the mRNA induction is stronger in females as well. The enkephalin gene induction by estrogen occurs rapidly, before lordosis behavior. Antisense reduction of enkephalin mRNA in the VMH reduces lordosis behavior, and molecular competition through thyroid hormone receptor systems blocks both enkephalin mRNA induction and lordosis behavior. The sexually differentiated aspects of these results are correlated with lordosis. Following from Chapters 3 and 4, we can say that multiple levels of evidence for the enkephalin gene show us the reliability of this topic in the developing field of neuroscience.

  Ethology

  When I entered neuroscience, two schools of thought competed with each other. Ethology, founded mainly by prominent European biologists, emphasized the noninvasive study of natural behavior patterns in natural environments. Biological psychology, imitating physics and mainly starting in the United States, emphasized controlled experimentation in the laboratory. We combined the two in our work. We chose a complete natural behavior that could be naturally evoked in the laboratory, and used neuroanatomical, electrophysiological, and molecular techniques to demonstrate its physical realization.

  Overdetermined

  Neural systems that govern reproductive behavior are not allowed to fail. To put it another way, species for which reproductive regulations were shaky would not last.

  In the service of rock-hard, reliable regulation, lordosis behavior performance is overdetermined in five ways. First, both with respect to lordosis and the ovulatory release of luteinizing hormone from the pituitary, progesterone massively amplifies the estrogen effect (reviewed in Fink, Pfaff, and Levine 2014). Thus, estrogens foster transcription of PR (Chapter 3), and progestins bind to PR, and, as a result, acting both in the VMH and in the pituitary, help to guarantee the timely performance of the reproductive behavior exactly at the time it is needed with respect to ovulation. This produces a strong and stable system.

  Second, not just one but several transcriptional systems (Chapter3) display a dynamic pattern. (a) Estrogen treatment increases transcription for that particular gene, and (b) the gene product facilitates lordosis behavior; therefore, (c) it is logical to infer that the particular transcriptional system represents one (of several) ways in which neuronal nuclear effects of estrogens drive the behavior. The results summarized in Chapter 3 show how the probability of reliable behavioral performance is increased by the several transcriptional systems’ separate but converging contributions.

  Thus, for example, as shown in Chapter 3, adrenergic inputs to VMH neurons would help to guarantee the effects of adequate central nervous system (CNS) arousal for sexual behaviors that require arousal. And the effect of enkephalins acting through δ-opioid receptors would help to guarantee sexual behavior performance under anxiety-producing environmental circumstances.

  Third, several neurotransmitter inputs to VMH neurons that cause increased electrical excitability of those neurons overdetermine their adequate electrical signaling to the lordosis behavioral circuit. I have mentioned adrenergic inputs acting through adrenergic α1 receptors, but muscarinic inputs, histaminergic inputs, and glutaminergic inputs are effective as well.

  Fourth, reliability and stability are shored up by coordinated activity among at least three non-neural organs in addition to nerve cell groups. That is, under the influence of the hypothalamus, 1) the anterior pituitary sends gonadotropins into the bloodstream in such a way that 2) cells in the ovary will produce estrogens (and later in the cycle progestins). Estrogens not only affect 3) the uterus (with its attendant innervation) but also feedback to the pituitary and the hypothalamus. The peak of estrogen levels, just before ovulation, is amplified by ovarian progesterone such that the ovulatory discharge of gonadotropins from the pituitary is synchronized with the performance of lordosis behavior.

  Fifth, though not emphasized here, I have written before about hormone-dependent behavioral funnels by which reproductively competent males and reproductively competent females can be brought to the same place at the same time (Pfaff 1999, 2010). Long ago, my laboratory illustrated this concept with data from hamsters. The male hamster’s testosterone-dependent flank marks (pheromones) attract the female hamster if and only if she has high estrogens. In turn, she deposits estrogen-dependent thick vaginal fluids (pheromones). These odors diffuse in time and space. Both animals go up the pheromone gradient, thus getting closer to each other. At that time, hormone-dependent ultrasounds provide signaling between male and female. Because ultrasounds do not go far (due to absorption on environmental surfaces) and are highly directional, the male and female can easily get together and then use cutaneous signals to direct behavior. Thus, hormone-dependent distance signals (or sexual readiness) lead the animals to get close for hormone-dependent proximal signals to occur, and they in turn to come in physical contact with each other for the male to mount and female’s lordosis to permit fertilization.

  Symmetry, Hierarchy

  Our lordosis behavior mechanisms are left-right symmetric, but modules along the anterior-posterior axis are strongly hierarchical. Hypothalamic hormone-dependent cells regulate the excitability of midbrain central grey neurons. In turn these neurons excite both lateral vestibulospinal and medial reticulospinal neurons, which in turn regulate the responses of lordosis-relevant motor neurons to behaviorally relevant sensory inputs. Thus, our neural network is modular and hierarchical.

  The accomplished ethologist Richard Dawkins (1976) sees great advantage in such an organization of mechanisms. In his chapter “Hierarchical Organization: A Candidate Principle for Ethology” he finds such systems to be efficient and stable. His argument fits
with my “Overdetermined” section above, as I talked about reliability and stability in lordosis behavior mechanisms.

  Universals Discovered

  In Chapter 1, I explained that the limbic-hypothalamic system of sex steroid-binding neurons discovered in the rat brain turned out to be universal among vertebrates. Likewise, gonadotropin-releasing hormone undergoes neuronal migration during development (Chapter 5) from the olfactory epithelium into the preoptic area / hypothalamus.

  In Chapter 7 there is a formulation of principles applicable to certain instinctive behavioral regulations. True for vertebrates in general, the principles deal with endocrine logic and the temporal and spatial features of hormone action on behavior. Most importantly, they say that neural mechanisms have been conserved to provide adaptive body / brain / behavior coordination in humans as in laboratory animals.

  Sex Differences

  Obviously, sex differences are prominent in the expression of reproductive behaviors (Pfaff 2010). What my laboratory showed (Chapter 9) was that testosterone levels just before and just after birth are critical for the sexual differentiation of behavior, and that the behavioral changes are substantial and permanent. Effectively we are considering the physiological side of libido (Pfaff 1999), analogous to Eric Kandel’s and Joseph LeDoux’s consideration of emotional states.

  In fact, Sonoko Ogawa’s study showed that we can use genetic manipulation to reverse behavioral sex roles (Ogawa et al. 1996). That said, when considering generalized CNS arousal (Pfaff 2006), I do not expect differences between sexes. The generalized drive that leads to sexual approaches should be the same in female and male.

  Emphasis on the Analyses of Function

  The approach to the vertebrate brain’s regulation of behavior embodied in this book and espoused for the future has two major features. First, the findings have proven extraordinarily durable; the discovery of hormone receptors in the brain was accomplished 50 years ago, and the ensuing findings have been heavily replicated. With modern methods we systematically followed a neo-Sherringtonian logic to work out the neural circuitry and then delve into detailed cell nuclear chemistry with the tools of molecular biology.

  Second, in working out the circuitry for a behavior our approach avoided a major mistake which is threatening current neuroscience: confusion about the definitions of nerve cell types. I recently read an article in Science magazine by Emily Underwood entitled “The Brain’s Identity Crisis: Will new tools for classifying neurons put a 150-year-long debate to rest?” Underwood was writing about the efforts of some otherwise very smart neuroscientists to classify interneurons. The meeting convened to try this took place, in a poetically correct choice, in the home town of the founder of modern neuroanatomy, the Spanish neuroanatomist Ramon y Cajal. Cajal used the silver nitrate stain discovered by the Italian neurologist Camillo Golgi to describe comprehensively many of the “types” of neurons in the brain. The “Identity Crisis” meeting was convened to define and name “neuronal cell types.” The meeting was a disaster, according to Underwood’s summary. A phrase used: “confusion was on full display.” In the words of Giorgio Ascoli from George Mason University “It was like the Tower of Babel. We thought we are all rational people, we all believe in data, we will create labels and codes and agree to them, and make the Babel tower issue go away. But by lunchtime on the first day, he says, it was clear that the task with which they were charged was, simply put, impossible.”

  These veteran neuroscientists were going about things the wrong way. For one thing, they were ignoring the subtle epigenetic signals imposing on individual neurons in the most intimate and singular way during development. That is, consider that Inna Tabansky, Joel N. H. Stern, and I (2016) attacked the fallacy of false equivalence among developing cells, for example, the mistaken assumption of identity among cells that look alike. As a highly skilled stem cell biologist and developmental biologist, Tabansky knew that starting with newly derived neurons that look alike, various ones among them would follow trajectories that would separate them into different functionally equivalent groups. Her view was not static, but dynamic. We quoted recent discoveries using single cell RNA sequencing, studies that revealed differences in gene expression among cells that might otherwise have been (mistakenly) considered identical. Thus, we envision the study of epigenetically influenced trajectories through space and time as the best way to characterize a neuron’s eventual functional identity.

  Tabansky et al (2016) is an essay in developmental neurobiology while, in parallel, this book demonstrates the success of approaching the vertebrate brain, likewise asking not “How does the brain work?” but instead asking: HOW is a specific FUNCTION of the central nervous system accomplished? Approaching the CNS in our way reveals the “identity” of individual neurons not a priori in the form of a dictionary sure to conflate neuronal apples with neuronal oranges, but instead it reveals each neuron’s type of specific contributions in action, as we show how a specific behavioral function is accomplished. That’s the best way to do it.

  Public Health

  It might be considered that, since sexual drives are connected to other instinctive behaviors such as aggression and parental behaviors, and since mechanisms for these primitive behaviors tend to be conserved (Pfaff, 1999, cf. LeDoux, 1996 and Kandel, 2012), understanding exactly how they work will be of use for application to questions in public health.

  Further Reading

  Dawkins, R. 1976. “Hierarchical Organization: A Candidate Principle for Ethology.” In Growing Points in Ethology, ed. P. Bateson and R. Hinde. Cambridge: Cambridge University Press, 7–54.

  Fink, G., D. W. Pfaff, and J. Levine. 2014. Handbook of Neuroendocrinology. San Diego: Academic.

  Kandel, E. R. 2012. The Age of Insight: The Quest to Understand the Unconscious in Art, Mind, and Brain, from Vienna 1900 to the Present. New York: Random House.

  LeDoux, J. 1996. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. New York: Simon & Schuster.

  Ogawa, S., J. Taylor, D. B. Lubahn, K. S. Korach, and D. W. Pfaff. 1996. “Reversal of Sex Roles in Genetic Female Mice by Disruption of Estrogen Receptor Gene.” Neuroendocrinology 64: 467–470.

  Pfaff, D. W. 1980. Estrogens and Brain Function. Heidelberg: Springer.

  ______. 1999. Drive: Neurobiological and Molecular Mechanisms of Sexual Motivation. Cambridge, MA: MIT Press.

  ______. 2006. Brain Arousal and Information Theory. Cambridge, MA: Harvard University Press.

  ______. 2010. Man and Woman: An Inside Story. New York: Oxford University Press.

  Tabansky, I., J. N. H. Stern, and D. W. Pfaff. 2015. “Implications of Epigenetic Variability within a Cell Population for ‘Cell Type’ Classification.” Frontiers in Behavioral Neuroscience 9: 342.

  Underwood, E. 2015. “The Brain’s Identity Crisis.” Science 349: 575–577.

  ACKNOWLEDGMENTS

  INDEX

  ACKNOWLEDGMENTS

  This book, which reflects on the parallel development of my career and of neuroscience, was over fifty years in the making. That is, it has been fifty years since my discovery of hormone receptors in the brain. If I were to thank everyone who has influenced this story of molecular biology built on neurophysiology built on neuroanatomy and behavior, these acknowledgments would be prohibitively long. So suffice it to say that no one’s career in science just happens without the support of a great many people. I am grateful to them all, and throughout the text I try to give credit wherever it is due.

  Nonetheless, a few people cannot go without mention here because they had a direct effect on the quality of this book. In this regard, I would first like to thank Dr. Sandra Sherman, who helped me clarify my language—at least to the extent that anything so complicated as neuroscience can be “clarified.” Her object was not just to make my presentation clear to the average reader, but to ensure that its intended readers could fully appreciate the development of neuroscience that I seek to describe. Sandra has helped me on previous boo
ks, but this one presented a unique challenge. How do you summarize fifty years of hard science, for a group of hard scientists, and still make sure that the main points do not fail to impress? To the degree that this book does justice to the field, I wish to thank Sandra for all of her help.

  I next wish to thank Janice Audet, editor at Harvard University Press. With her perspicacity, Janice understood at once the importance of telling a story of fifty years during the development of this exciting field and making this book valuable to researchers across a rather broad intellectual spectrum. Also appreciated are the anonymous reviewers whose suggestions I have followed, to the benefit of the final manuscript.

  Finally, I would like to thank the scientists formerly in my laboratory with whom I have spent so much time over the years. You will frequently see the name of Lee-Ming Kow, who still works with me in my laboratory at Rockefeller, and Sonoko Ogawa, who is now a professor at Tsukuba University in Japan. The wonderful research teams we have formed over the years account for the crescendo of scientific proofs—from hormone receptors in the brain, to neuroanatomy, to neurophysiology, to molecular biology—to offer a unified story of how the vertebrate brain regulates behavior. The creative atmosphere at the Rockefeller University allows this kind of ambitious research program to be initiated, maintained, and brought to fruition.

  INDEX

  Alpha-adrenergic receptors, 89–91

  Amygdala: androgen-concentrating neurons in, 22; connections to, 35; ER-α and OTR transcription in, 137; ER-α and OTR transcription in VMH and POA, 223–224; ER-α gene in medial (MePD), 133–134; estrogen-binding neurons in medial nucleus, 13–14, 17, 19, 20–21, 27; limbic-hypothalamic system, 97; posterodorsal (MePDA), 135–136; vlVMH projections to, 42