How the Vertebrate Brain Regulates Behavior

Home > Other > How the Vertebrate Brain Regulates Behavior > Page 16
How the Vertebrate Brain Regulates Behavior Page 16

by Donald Pfaff

We investigated the role of gene expression of the estrogen receptor α form (ER-α) in the regulation of female reproductive behavior by using ER-α knockout (αERKO) mice, females deficient specifically for the ER-α but not the ER-β gene (Ogawa, Eng, et al. 1998). Estrogen-treated, or estrogen plus progesterone–treated αERKO mice did not show any lordosis response under any experimental conditions (Figure 4.1). That is, even with maximal hormonal facilitation, αERKO females could not show lordosis behavior.

  Figure 4.1. Microinjection of a viral vector encoding a small-interfering RNA that blocks expression of the estrogen receptor-α gene (AAV.H1.ER1) into ventromedial hypothalamus (A) abolishes courtship behaviors, (B) abolishes lordosis behavior, and (C) increases rejection of the male. (Adapted from Ogawa, Eng, et al. 1998.)

  Detailed behavioral analysis revealed that αERKO females were also deficient in the sexual behavioral interactions that usually precede the lordosis response. They were extremely rejective toward attempted mounts by stud male mice, which could not show any intromissions. During resident–intruder aggression tests, gonadally intact αERKO females were more aggressive toward female intruder mice than were the wild-type (WT) mice. Taken together, a long set of experiments showed that ER-α gene expression plays a key role in female mice, most prominently for sexual behavior but also for other interrelated behaviors such as parental and aggressive behaviors. Emilie Rissman’s laboratory at the University of Virginia replicated our results. She and others have worked for years to turn this set of findings into an entire field of neuroendocrinology.

  In the most striking experiment, Sonoko Ogawa (1996) put αERKO females in cages with resident stud males; when the females were mounted by the males, the lordosis behavior was much reduced for at least two reasons: less responsiveness to somatosensory stimuli on the hindquarters of the female (which ordinarily would lead to lordosis behavior), and the αERKO females being treated as intruder males by the resident stud males and thus being attacked. When mounted, the αERKO females showed strong rejection behavior, including kicking the male with the hind legs and rapid flight. In addition, aggression by αERKO females toward other females was significantly increased, and maternal-like retrieving was reduced. Thus, disruption of the ER-α gene led these females to lose their normal female-typical behavior and to behave and to be treated more like males. That is, imposing a knockout of the ER-α gene led to a reversal of sex roles.

  Our approaches to genetic manipulations in the brain were vastly expanded when Michael Kaplitt came to the laboratory in 1988. A virologist, Michael had training in Tom Schenck’s laboratory at Princeton; being a highly articulate and enthusiastic student, he quickly convinced me that we would be able to use viruses to modify gene expression in the brain. Using a replication-deficient herpes simplex virus, he quickly achieved the first use of a virus to express a foreign gene in a mammalian brain (Kaplitt et al. 1991). Then, we published on the first use of adeno-associated virus (AAV) as a gene delivery vehicle for the brain (Kaplitt, Leone, et al. 1994; Kaplitt, Kwong, et al. 1994).

  Just as important, he made us comfortable with the design and use of viral vectors, which Sonoko Ogawa and I would use for years. For example, short hairpin RNA (shRNA) are artificial RNA molecules with a tight hairpin turn that can be used to silence target gene expression via RNA interference, with a relatively low rate of degradation and turnover. Both vectors (the control AAV-luciferase and the experimental AAV-ER-α) contain, in addition, an independent enhanced green fluorescent protein (EGFP) expression cassette under the control of a hybrid cytomegalovirus / chicken β-actin promotor. The control AAV-luciferase was used to control for any potential nonspecific adverse effects of surgery or toxicity of encoded products, and the EGFP was used as a reporter to visualize transduced neurons.

  Michael’s assistant, Dr. Sergei Musatov, would provide these reagents that would allow us to discover the necessity of patterns of gene expression in specific nerve cell groups such as the ventromedial nucleus of the hypothalamus (VMH) for the proper regulation of instinctive behaviors.

  Thus, Sergei Musatov started his work with the certainty that ER-α plays a major role in the regulation of neuroendocrine functions and behaviors by estrogens (Musatov et al. 2006). Although the generation of ER-α knockout mice had advanced our knowledge of ER-α functions, gene deletion using this method is global and potentially confounded by developmental consequences. To achieve a site-specific knockdown of ER-α in the normally developed adult brain, we generated an AAV vector expressing a shRNA targeting ER-α. After bilateral injection of this vector into the VMH in female mice, the expression levels of ER-α as well as the estrogen-inducible progesterone receptor (PR) were profoundly reduced despite the continued presence of this receptor elsewhere in the brain. Thus, our manipulation was validated. Functionally, silencing of ER-α in the VMH abolished female lordosis behavior and courtship behaviors while enhancing rejection behaviors by the female. These AAV-mediated long-term knockdown of genes can be used to delineate their effects on complex behaviors in a specific nerve cell group.

  Working with my imaginative friend, the Norwegian professor Anders Ågmo, we invented “seminatural environments” to extend my earlier findings (Snoeren et al. 2015). We investigated the role of ER-α in the VMH, the preoptic area (POA), the medial amygdala (MePD), and the bed nucleus of stria terminalis (BNST) in sociosexual behavior in female rats. The idea was to study behavior precisely in a larger and more complicated environment than is usually done. The rats were housed in groups in the seminatural environment for 8 days. A group of rats consisted of four females and three males, and all were unfamiliar with each other and sexually inexperienced. Before each group was introduced, the floor in the open area, the tunnels, and the nest boxes were covered with approximately 2 cm of aspen wood chips. Approximately 2 kg of food pellets were put on the floor, close to a corner in the open area. Twelve aspen wood sticks were randomly distributed in the open area, and three red polycarbonate huts were irregularly placed closed to the middle. In addition, six pieces of a small square mat of nonwoven hemp fibers were put in each nest box in the burrow area. To distinguish among the rats on the video record, rectangles were carefully shaved on the back of the rats the day of experimentation.

  We performed two sets of experiments: the VMH and POA were investigated in the first set, and the MePD and BNST in the second set. In the first, we used a shRNA encoded within an AAV vector directed against the ERα gene to reduce the expression of ERα in the VMH or POA. As mentioned, in comparison to traditional test setups, the seminatural environment provides an arena in which the rats can express their full behavioral repertoire, which allowed us to investigate multiple aspects of social and sexual behavior in groups of rats. As expected, a reduction of ERα expression in the VMH or POA diminished the display of paracopulatory behaviors. This suggests that ERα gene expression in the VMH and POA plays an important role in intrinsic sexual motivation. The reduction in ERα did not affect the general social behavior of the females, suggesting some behavioral specificity to the effect. In the second experiment, the expression of ERα in the MePD and BNST, on the other hand, played no role in these sociosexual behaviors, indicating neuroanatomical specificity of our results.

  The kinds of locomotion used in courtship behaviors by the females were also of interest (Ogawa et al. 2003). For example, estrogens are known to increase running wheel activity of rodents primarily by acting on the medial preoptic area (mPOA). The mechanisms of this estrogenic regulation of running wheel activity are not completely understood. In particular, little is known about the separate roles of two types of estrogen receptors, ER-α and ER-β, both of which are expressed in mPOA neurons. In this study the effects of continuous estrogen treatment on running wheel activity were examined in male and female mice specifically lacking either the αERKO or the ER-β (βERKO) gene.

  The mice were gonadectomized and later implanted with either an estradiol benzoate (EB) or with a placebo control pellet. The same mice we
re also tested for open field activity before and after EB implants. In both female and male αERKO mice, the running wheel activity was not different from that in the corresponding wild-type (αWT) mice in placebo control groups. But in both females and males it was increased by EB only in αWT, not αERKO, mice. This was the main result: the type of locomotor activity used in courtship behaviors could not be increased by estrogens when the gene for ER-α was knocked out.

  This result just described was gene specific. In βERKO mice both doses of EB equally increased running wheel activity in both sexes just as they did in βWT mice. Absolute numbers of daily revolutions of EB-treated groups, however, were significantly lower in βERKO females compared with βWT females. Before EB treatment, αERKO female were significantly less active than αWT mice in open field tests, whereas βERKO females tended to be more active than βWT mice. This same pattern did not show up in the open field, indicating some behavioral specificity for these results.

  The locomotor activity that we measured is necessarily brought into play during the search for and approach to a potential partner. An shRNA encoded, as described previously, within an AAV vector directed against the ER-α gene (or containing a nonsense base sequence as a control treatment) was injected bilaterally into the VMH or the posterodorsal amygdala (MePDA) of female rats (Spiteri et al. 2010). After an 80 percent reduction of the expression of ER-α in the VMH, sexual approaches to the male were totally absent even after treatment with estradiol and progesterone. We also replicated the loss of lordosis, mentioned earlier in Sonoko Ogawa’s work, when ER-α expression was suppressed. Suppression of the ER-α in the MePDA lacked these effects, thus showing the neuroanatomical specificity of the VMH result.

  At least part of the requirement for a normal ER-α gene comes from the necessity for ER-α expression in the mPOA. Cooperating again with my Norwegian friend Anders Ågmo, we blocked ER-α expression in the POA specifically, and showed that females with reduced preoptic ER-α expression failed to show enhanced locomotor activity after treatment with EB (Spiteri et al. 2012).

  What about the nutritional requirements that go along with reproductive cycles? ER-α plays a pivotal role in the regulation of food intake and energy expenditure by estrogens. Although we had documented that a disruption of ER-α signaling in αERKO mice leads to an obese phenotype (Geary et al. 2001), the sites of estrogen action and mechanisms underlying this phenomenon had to be worked out. All we knew was that αERKO worked and βERKO had no effect. We then exploited RNA interference mediated by the AAV vectors mentioned previously to achieve focused silencing of ER-α in the VMH, a key center of energy homeostasis (Musatov et al. 2007). After ER-α expression in the VMH had been suppressed, female mice and rats developed a phenotype characteristic for the metabolic syndrome, marked by obesity, hyperphagia, impaired tolerance to glucose, and reduced energy expenditure. This phenotype persisted despite normal ER-α levels elsewhere in the brain. We think that VMH neurons integrate metabolic regulation appropriate for reproduction with the actual performance of reproductive behavior.

  Elena Choleris, now a professor at the University of Guelph in Ontario, Canada, was most interested in the processes of social recognition by the female that go along with successful reproduction (Choleris et al. 2003). That is, estrogens regulate many physiological and behavioral processes, some of which are connected to reproduction. These include a variety of social behaviors. We implicated no less than four gene products in a “micronet” required for mammalian social recognition, through which an individual learns to recognize other individuals. Female mice whose genes for the neuropeptide oxytocin (OT) or the ER-β or ER-α had been selectively knocked out were deficient specifically in social recognition and social anxiety. There was a remarkable parallelism among results from these three separate gene knockouts.

  The data, taken together, showed the involvement in social recognition of the four genes coding for ER-α, ER-β, OT, and the OT receptor (OTR). We thus proposed the four-gene micronet, which links hypothalamic and limbic forebrain neurons in the estrogen control over the OT regulation of social recognition. In our model, estrogens act on the OT system at two levels: through ER-β they regulate the production of OT in the hypothalamic paraventricular nucleus, and through ER-α they drive the transcription of the OTR in the amygdala. The proper operation of a social recognition mechanism allows for the expression of appropriate social behaviors, be they aggressive or sexual.

  To confirm and extend that extensive set of experiments, we followed them up with a binary social discrimination assay, in which the animals are given a simultaneous choice between a familiar and a previously unknown individual (Choleris et al. 2006), thus offering a more direct test of social recognition than in the 2003 study. The 2003 results were confirmed. Differently from their WT controls, when given a choice the knockout mice showed either reduced (βERKO) or completely impaired (OTKO and αERKO) social discrimination. Detailed behavioral analyses indicated that all of the knockout mice had reduced anxiety-related stretched approaches to the social stimulus, with no overall impairment in horizontal and vertical activity, nonsocial investigation, and various other behaviors such as self-grooming, digging, and inactivity. Thus, the gene knockout results were specific to the social domain. Our four-gene micronet model for social recognition still stands.

  Maternal behavior is still another crucial behavior in the chain of behaviors needed for successful reproduction. We used small interfering RNA (siRNA) silencing of gene expression in neurons of the POA (Ribeiro et al. 2012). The mPOA has been shown to be intricately involved in many behaviors, including locomotion, sexual behavior, maternal care, and aggression. The gene encoding ER-α protein is expressed in POA neurons (Chapter 1), and a very dense immunoreactive field of ER-α is found in the preoptic region. ER-α knockout animals show deficits in maternal care (Figure 4.2) and sexual behavior, and fail to exhibit increases in these behaviors in response to systemic estradiol treatment. In the 2012 study, we used viral-vector–mediated RNA interference to silence ER-α expression specifically in the POA of female mice and measured a variety of behaviors, including social and sexual aggression, maternal care, and arousal activity. The massive reduction in the expression of ER-α in preoptic neurons was validated.

  Figure 4.2. Microinjection of a viral vector encoding a small-interfering RNA that blocks expression of the estrogen receptor-alpha gene (AAV.H1.ER1) into medial preoptic area causes failure of maternal behavior. (A) Lack of licking and nursing the pups. (B) Increased latency to pup retrieval. *, p<0.01 ***. [<0.001. (Adapted from Ribeiro et al. 2012.)

  Suppression of ER-α expression in the POA specifically almost completely abolished maternal care (95 percent reduction), significantly increasing the latency to pup retrieval and significantly reducing the time that the mothers spent nursing and licking the pups. Strikingly, maternal aggression toward a male intruder was not different between the control and preoptic ER-α–silenced mice, demonstrating the remarkably specific role of ER-α in these neurons. Reduction of ER-α expression in preoptic neurons significantly decreased sexual behavior in female mice and increased aggression toward both sexual partners and male intruders in a seminatural environment. Estrogen-dependent increases in arousal, measured by home cage activity, were not mediated by ER-α expression in the preoptic neurons we targeted, as ER-α–suppressed mice had increases similar to the control mice. Thus, we have established that a specific gene in a specific group of neurons is required, specifically, for a crucially important natural behavior.

  Thus, the entire chain of behaviors essential for the reproductive process—locomotion and “courtship” by the female, lordosis behavior, nutritional balance, and maternal behaviors—all depend on regulated gene expression of ER-α. These findings showed that we could go from applied molecular biology (Chapter 3) to entire sets of mammalian behavior. Such powerful gene / behavior causal relations form the reliable foundation of a field of work, and they are integrated with body-wide reproductive
physiology in Chapter 7 and are treated conceptually in Chapter 10.

  The Gene for Estrogen Receptor β

  The most striking single result was, as mentioned earlier, that knocking out ER-β not only did not lead to an absence of lordosis behavior, but led to βERKO females showing slightly higher levels of reproductive behavior—in dramatic contrast to ER-α. This result surprised us a great deal because of the strong phenotype of ER-β in brain tissue, as discovered by Jan-Ake Gustafsson at the Karolinska Institutet. The Gustafsson laboratory and others have been highly productive in demonstrating ER-β’s role in brain development, synaptic plasticity, and neuroimmune phenomena. Initially studied with ER-β knockout mice, now ER-β–selective ligands are used to advantage. Gustafsson has thus elucidated ER-β functions essential for maintaining the extracellular matrix, cell survival, and inflammatory processes.

  Accordingly, lordosis behavior appears stronger than normal in βERKO mice (Ogawa et al. 1999). Looking back, we had already shown that the lack of a functional ER-α gene greatly affects reproduction-related behaviors in female mice. However, the widespread expression in the brain of Gustafsson’s novel ER-β mice, demanded that we also examine the possible participation of ER-β in regulation of these behaviors. In dramatic contrast to our results with αERKO males, βERKO males performed at least as well as WT controls in sexual behavior tests. Moreover, not only did βERKO males exhibit normal male-typical aggressive behavior, including offensive attacks, but they also showed higher levels of aggression than WT mice under certain conditions of social experience. These data revealed a significant interaction between genotype and social experience with respect to aggressive behavior. Finally, females lacking a functional β isoform of the ER gene showed normal lordosis and courtship behaviors, extending in some cases beyond the day of behavioral estrus. These results highlight the importance of ER-α for the normal expression of natural reproductive behaviors in both sexes and also provide a background for future studies evaluating ER-β gene contributions to other, nonreproductive behaviors.

 

‹ Prev