How the Vertebrate Brain Regulates Behavior

by Donald Pfaff

  Schwanzel-Fukuda, M., and D. W. Pfaff. 1989. “Origin of Luteinizing Hormone-Releasing Hormone Neurons.” Nature 338: 161–164.

  Snoeren, E. M., E. Antonio-Cabrera, T. Spiteri, S. Musatov, S. Ogawa, D. W. Pfaff, and A. Ågmo. 2015. “Role of Oestrogen α Receptors in Sociosexual Behavior in Female Rats Housed in a Seminatural Environment.” Journal of Neuroendocrinology 27: 803–818.

  Spiteri, T., S. Musatov, S. Ogawa, A. Ribeiro, D. W. Pfaff, and A. Ågmo. 2010. “Estrogen-Induced Sexual Incentive Motivation, Proceptivity and Receptivity Depend on a Functional Estrogen Receptor Alpha in the Ventromedial Nucleus of the Hypothalamus but Not in the Amygdala.” Neuroendocrinology 91: 142–154.

  Spiteri, T., S. Ogawa, S. Musatov, D. W. Pfaff, and A. Ågmo. 2012. “The Role of the Estrogen Receptor α in the Medial Preoptic Area in Sexual Incentive Motivation, Proceptivity and Receptivity, Anxiety, and Wheel Running in Female Rats.” Behavioural Brain Research 230: 11–20.

  Vasudevan, N., M. Morgan, D. W. Pfaff, and S. Ogawa. 2013. “Distinct Behavioral Phenotypes in Male Mice Lacking the Thyroid Hormone Receptor α1 or β Isoforms. Hormones and Behavior 63: 742–751.

  5

  NEUROPEPTIDE: GONADOTROPIN-RELEASING HORMONE

  Problem: I wanted to account for the integration of a social behavior, lordosis, with whole-body preparations for reproduction, by studying gonadotropin-releasing hormone, a decapeptide. Thus, this work resulted in causal links: chemistry to physiology to behavior.

  My discovery of the migration during development of gonadotropin-releasing hormone (GnRH) neurons from the olfactory epithelium into the brain depended on the experience and influence of two “artists” of much different sorts. By contrast, when I got the idea that I could use GnRH to drive behavior it was a lightning-fast idea, and the wide acceptance of my behavioral data was assisted by the work of a friendly competitor. However, first came the demonstration that GnRH neurons are not born in the brain, as are other neurons. And again, you’ll remember that the older term for GnRH was luteinizing hormone-releasing hormone (LHRH).

  Development

  One of the two artists, Marlene Schwanzel-Fukuda, had a long career as a medical illustrator before she got her doctorate in neuroanatomy and joined my laboratory. With the close collaboration of Joan Morrell at Rockefeller University, Marlene was doing straightforward immunocytochemical studies of GnRH neurons. For example, Marlene got interested in LHRH neurons associated with a little-studied input to the forebrain, the nervus terminalis (Schwanzel-Fukuda et al. 1987).

  One day, looking at her data on a developmental time series of GnRH immunoreactive neurons, I saw a pattern of results that screamed “migration.” Why? The answer lay in the histochemical artistry of my Ph.D. advisor at the Massachusetts Institute of Technology (MIT), Joseph Altman.

  Joe, an immigrant from Hungary, had distinguished himself at New York University as a neuroscience student and had come to MIT as a junior faculty member. Joe was a different kind of artist than Marlene Schwanzel-Fukuda—he used precise histochemical methodology to produce stunningly beautiful neuroanatomical illustrated results. Before I got to MIT as a graduate student, he had discovered postnatal neurogenesis; when I got there in 1961 he had me helping him trace these newborn neurons in the central nervous system. Of course, as related in Chapter 1, I took the initiative to study and discover hormone receptors in the brain, but my early experience with Joe’s work, tracing the movements of identified cells, primed my mind to see migrations of neurons.

  Here I sketch some of the details of Marlene Schwanzel-Fukuda’s demonstration of the startlingly novel GnRH neuronal migration during development (Schwanzel-Fukuda and Pfaff 1989). We began with the knowledge that in the adult these neurons are found in the septal-preoptic nuclei and in the hypothalamus, that they control the release of gonadotropic hormones from the anterior pituitary gland, and that they facilitate reproductive behavior. Marlene got interested in the fact that GnRH-expressing neurons are also found in the nervus terminalis, a cranial nerve that is a part of the accessory olfactory system and projects directly from the nose to the septal-preoptic nuclei in the brain. During development, GnRH immunoreactivity is detected in the peripheral parts of the nervus terminalis before it is found in the brain. Using a combination of LHRH immunocytochemistry and tritiated thymidine autoradiography in fetal mice, we showed that LHRH neurons originate in the medial olfactory placode of the developing nose, migrate across the nasal septum, and enter the forebrain with the nervus terminalis, arching into the septal-preoptic area and hypothalamus (Figure 5.1). Clinically, this migratory route for LHRH-expressing neurons could explain the deficiency of gonadotropins seen in Kallmann’s syndrome (hypogonadotropic hypogonadism with anosmia).

  Marlene and I made the migration discovery in mice, and subsequent work in my own and other laboratories showed that the GnRH neuronal migration from nose into brain is a universal vertebrate phenomenon. Here I will emphasize the extremes of neural system sophistication among vertebrates, specifically fish and humans.

  Fish

  Ishwar Parhar, a fish brain biologist, received his doctorate from the National University of Singapore and was determined to work with us. I told him repeatedly that we did not have any money for his salary, but he was and is such a highly motivated scholar that he insisted on coming anyway. For the first 6 months he lived in an underused room in our large laboratory, a room that we designated “hotel hypothalamus.” Then he got a paying job in Torsten Wiesl’s laboratory before getting his own fellowship. His first experiment used immunocytochemistry and in situ hybridization (Parhar, Pfaff, and Schwanzel-Fukuda 1995) to test the hypothesis that GnRH neurons are formed in the olfactory placode during embryonic development in a salmonid fish, Oncorhynchus nerka.

  We examined the development of GnRH neurons and the pituitary cell types from 19 through 910 days after fertilization. Immunoreactive GnRH was first detected at 19 days in the cells of the olfactory placode. GnRH immunoreactivity was not detected in any other structure of the central nervous system at this age. By day 24, GnRH-immunoreactive neurons were seen in the apical, intermediate, and basal layers of the olfactory placode. From days 30 through 51, GnRH neurons were seen emerging from the epithelium, along the olfactory nerve, and at the rostral olfactory bulb. By day 41, GnRH immunoreactivity was lost in the nasal epithelium.

  In the 72-day-old fish, most of the GnRH neuronal population was found in ganglia of the nervus terminalis, at the cribriform bone (gCB), and at the rostral olfactory bulb (gROB). On day 293, a decrease in GnRH-immunoreactive neurons in the gCB and gROB was concomitant with an initial appearance of GnRH-immunoreactive neurons and fibers along the caudoventral olfactory bulb. By day 462, the distribution of GnRH neurons and fibers was almost similar to adults. In maturing adults (910 days), GnRH-immunoreactive neurons were rarely seen in the nasal regions and were primarily found in the basal forebrain. GnRH fibers were widespread in the brain, proximal para distalis, and in the pars intermedia of the pituitary. Thus, our first fish study (Figure 5.2) supported the generality of Marlene’s discovery that neurons expressing GnRH messenger RNA (mRNA) and peptide originate in the medial olfactory placode and migrate into the basal forebrain during development.

  To follow this up, Ishwar studied a major food fish from Southeast Asia, the tilapia. Thus, expression of multiple molecular forms of GnRH mRNAs and GnRH peptides were examined in the brains of Oreochromis mossambicus as well as sockeye salmon (Oncorhynchus nerka), using in situ hybridization histochemistry and immunohistochemical techniques (Parhar, Pfaff, and Schwanzel-Fukuda 1996). From a neuroanatomical point of view, there was a good correlation between the distributions of GnRH mRNA and cells containing GnRH peptide. Although the brains of tilapia and the sockeye were immunoreactive to three forms of the GnRH molecule (salmon, mammal, chicken-II), the GnRH mRNA expression was site specific and species specific. In the tilapia, ganglionic cells of the nucleus olfactoretinalis, basal telencephalon, and the anteroventral preoptic ar
ea were immunoreactive to salmon and mammalian GnRH peptide. Neurons of the nucleus olfactoretinalis expressed cichlid-GnRH I mRNA. Midbrain neurons were immunoreactive to salmon-GnRH but expressed cichlid-GnRH II beta (= chicken-GnRH II) mRNA hybridization signals. In the sockeye, ganglionic cells along the extracerebral course of the nervus terminalis were immunoreactive to mammalian, chicken II, and salmon GnRH. These neurons expressed only salmon GnRH mRNA hybridization signals. Intracerebral GnRH expression in the sockeye was delayed till smoltification. Overall, Marlene Schwanzel-Fukuda’s discovery of a migration was supported again.

  Figure 5.1. Top: Sagittal sections through developing mouse head, looking from the right side. The black dots indicate gonadotropin-releasing hormone (GnRH) cells migrating from the olfactory placode into the brain on embryonic days 11 through 16. Bottom: Histogram and graph quantifying the GnRH neuronal position to document the migration from olfactory placode into brain. (Adapted from Schwanzel-Fukuda and Pfaff 1989.)

  Figure 5.2. Gonadotropin-releasing hormone (GnRH) neural migration in the embryonic and larval stages of salmon. On embryonic day 25, salmon GnRH type 3 neurons are seen in the olfactory placode. At 30 days after fertilization, GnRH3 neurons appear at the basal regions of the olfactory epithelium (OE). At 40 days after fertilization, GnRH neurons (arrows) are seen along the olfactory nerve (ON) to the terminal nerve in the forebrain (FB), including the preoptic area (POA). The dotted box on left is shown in the photomicrograph on right. (Adapted from Parhar, Pfaff, and Schwanzel-Fukuda 1995.)

  Human Brain

  In 1989 after Nature had published Marlene’s discovery in the mouse brain (Schwanzel-Fukuda and Pfaff 1989), we had the opportunity to use human tissue. This is what we had in mind: Kallmann syndrome (inherited hypogonadotropic hypogonadism with anosmia) is associated with an X-chromosome deletion at Xp 22.3. In a Kallmann fetus, we found an absence of cells expressing luteinizing hormone-releasing hormone (GnRH) in the brain despite dense clusters of GnRH cells and fibers in the nose. Cells and neurites containing GnRH end in a tangle beneath the forebrain, within the dural layers of the meninges, on the dorsal surface of the cribriform plate of the ethmoid bone. Apparently, the GnRH neurons had gone through their final cell division in the medial extent of the olfactory placode, but at the beginning of their migration they got “stuck” in a neurovascular structure at the cribriform plate. Normal fetal brains, matched for age and sex, had GnRH cells and fibers, as expected, in the hypothalamus and preoptic area. Because cells expressing GnRH recently were discovered to migrate from the olfactory placode into the brain, it appears that the hypogonadotropism of Kallmann syndrome can be accounted for by a failure of GnRH cells to migrate into the brain (Schwanzel-Fukuda, Bick, and Pfaff 1989).

  Our neuroanatomical work in New York was complementary to the genetic work of Christine Petit in Paris. She demonstrated the Kall gene mutation that gave rise to our results. Thus, in humans an interruption of the GnRH neuronal migration, especially in men, causes a state in which the body does not produce adequate amounts of the sex hormone testosterone. This hypogonadal state is associated with loss of libido. In Chapter 4, I recounted how this state of affairs comprises a solid causal relation of genes onto human behavior, but one that requires understanding a chain of six causal steps. Understandably, Marlene Schwanzel-Fukuda and I have been extremely pleased to see that the field has revealed our migration discovery as universal among vertebrates.

  Our study published in Nature was quickly replicated. Susan Wray, a fine developmental neuroanatomist working at the National Institutes of Health, was collecting data on GnRH neurons. At the Society for Neuroscience meetings, Marlene’s posters usually gathered quite a crowd. As an artist, her posters were famous for their designs and for the brilliance of her neuroanatomical illustrations. My job regarding these posters was to patrol the periphery of the crowd and explain Marlene’s findings to important colleagues who otherwise might wander away. Susan Wray came by and from the edge of the onlookers caught sight of Marlene’s demonstration of a GnRH neuronal migration. Her eyes suddenly sparkled, her face lit up, and she started writing fast in her notebook. About 10 months later her report on replicating our work appeared.

  The cell biology and neuroanatomy of GnRH (also known as LHRH) neurons have been studied extensively (in our laboratory, e.g., Schwanzel-Fukuda, Jorgenson, et al., 1992; Schwanzel-Fukuda, Zheng, et al. 1992; Zheng, Pfaff, and Schwanzel-Fukuda 1992). But we knew the mechanisms would be complicated. Our first thought was to investigate the roles of cell adhesion molecules, a subject I knew nothing about. So I swallowed my pride to go downstairs in my building at Rockefeller and ask the expert, Professor Gerald Edelman. This was difficult for me to do because Gerry was known as a supercilious, if brilliant, scientist. A Nobel Prize–winner for immunology, it was said, snidely, of Gerry “that he was working on his second Nobel Prize, in neurobiology.” But during that conversation, Gerry was extremely nice to me in three separate but related ways. First, he gave me good advice about neural cell adhesion molecules (NCAM). Second, he also gave me good reagents, such as antibodies. Third, he permitted us to work collaboratively with one of his brightest laboratory members, Katherine Crossin. In the subsequent experiments we made three interesting advances.

  First, in 1992, we reported that contact between the developing forebrain and the ingrowing central processes of the olfactory, vomeronasal, and terminal nerves is preceded by a migration of NCAM immunoreactive cells from the epithelium of the olfactory pit and the formation of an NCAM immunoreactive cellular aggregate in the mesenchyme located between the olfactory pit and the forebrain. The axons of the olfactory, vomeronasal, and terminal nerves, also NCAM immunoreactive, grow into the cellular aggregate, which as development proceeds, becomes continuous with the rostral tip of the forebrain. The lateral and more rostral part of the cellular aggregate receives the ingrowing axons of the olfactory nerves and becomes the olfactory nerve layer of the olfactory bulb. The medial, more caudal part receives the central processes of the vomeronasal and terminal nerves. The vomeronasal nerve ends in the accessory olfactory bulb. The central processes of the terminal nerve end in the medial forebrain. GnRH immunoreactive neurons, like the vomeronasal and terminal nerves, originate from the medial part of the olfactory pit. These GnRH cells migrate into the brain along and within a scaffolding formed by the NCAM immunoreactive axons of the vomeronasal and terminal nerves, and they are never seen independent of this NCAM scaffold as they cross the nasal lamina propria.

  The results suggest that 1) NCAM is likely to be necessary for scaffold formation and 2) the scaffold may be essential for the subsequent migration of GnRH neurons into the brain. Because they aggregate, the migrating GnRH immunoreactive neurons on which we did not detect NCAM immunoreactivity may interact via other cell adhesion molecules. Inasmuch as the interaction between the GnRH immunoreactive neurons and the NCAM immunoreactive scaffold is heterotypic, the possibility of a heterophilic (NCAM to other cell adhesion molecules) interaction was not ruled out. These findings focused our attention on the functional role of NCAM in this migratory system.

  Second, in 1994, we followed up that functional implication (Schwanzel-Fukuda et al. 1994). The neurons that synthesize and release luteinizing hormone-releasing hormone (GnRH) were known to originate in the epithelium of the medial olfactory pit and to migrate into the brain along a scaffolding made up of NCAM immunoreactive branches of the terminal and vomeronasal nerves. These GnRH neurons, studied by immunocytochemical and autoradiographic procedures, were found to originate in mice within a very short period of embryogenesis (specifically day 10) and to follow a remarkably ordered spatiotemporal course along the migration route into the brain. The purpose of the new experiments was to determine whether perturbation of the NCAM immunoreactive migration route at a particular time in development would arrest the migration of GnRH neurons into the brain.

  We found that a 1-microliter injection of antiserum to NCAM into the area of the olfactory pit o
n day 10 of embryogenesis significantly reduced the number of GnRH immunoreactive neurons seen in the epithelium of the medial olfactory pit, with a concomitant significant reduction in the number of GnRH immunoreactive cells seen outside of the placode on the migration route. These results confirmed our initial hypothesis that GnRH neurons migrate from the epithelium of the olfactory pit to the brain and indicate that NCAM plays a causal role in this phenomenon.

  Human brain, with NCAM. Our third advance came in 1996, when Marlene had the opportunity to put our NCAM knowledge to work with human tissue (Schwanzel-Fukuda et al. 1996). GnRH neurons originate in the epithelium of the medial olfactory pit and migrate from the nose into the forebrain along nerve fibers rich in NCAM. This third study examined the ontogenesis of GnRH neurons in early human embryos and found a similar pattern of development of these cells. GnRH immunoreactivity was detected in the epithelium of the medial olfactory pit and in cells associated with the terminal-vomeronasal nerves at 42 (but not 28 to 32) days of gestation. The migration route of these cells was examined with antibodies to NCAM and antibodies to polysialic acid (PSA-NCAM), which is present on NCAM at certain stages of development. NCAM immunoreactivity was present in a population of cells in the olfactory placode of the earliest embryos examined (28–32 days) and later (42 and 46 days) throughout the migration route. PSA-NCAM immunoreactivity was not detected until 42 days and was present in a more limited distribution in nerve fibers streaming from the olfactory placode and along the caudal part of the migration route below the forebrain. Previous studies indicated that the highly sialylated form of NCAM is less adhesive. The PSA-NCAM may therefore facilitate the migration of these cells by lessening the adhesion between the fascicles that make up the migration route, expediting the passage of cords of GnRH cells between the nerve fibers as these cells move toward the brain.