Saturday, May 17, 2014

Evolving the Future: Toward a Science of Intentional Change

From the journal Behavioral and Brain Science, this is a very interesting and long article on the process of creating intentional change in people. Two of the authors are the evolutionary sociobiologist, David Sloan Wilson, and the creator of Acceptance and Commitment Therapy (ACT), Steven C. Hayes. I am only including the abstract and introduction, along with the first section - the whole paper is 91 pages.

Full Citation:
Wilson, DS, Hayes, SC, Biglan, A, Embry, DD. (2014, May 15). Evolving the Future: Toward a Science of Intentional Change. Behavioral and Brain Sciences; pp 1-99. DOI: 

Evolving the Future: Toward a Science of Intentional Change

David Sloan Wilson [a1 c1], Steven C. Hayes [a2], Anthony Biglan [a3] and Dennis D. Embry [a4]
a1. SUNY Distinguished Professor, Departments of Biology and Anthropology, Binghamton University, Binghamton, NY 13903. Email: Home page:
a2. Foundation Professor, Department of Psychology, University of Nevada, Reno, NV 89557-0062. Email: Home page:
a3. Senior Scientist, Oregon Research Institute, 1715 Franklin Boulevard, Eugene, OR 97403. Email: Home page:
a4. CEO, PAXIS, Inc. Tucson, Arizona. Email: Home page:


Humans possess great capacity for behavioral and cultural change, but our ability to manage change is still limited. This article has two major objectives: first, to sketch a basic science of intentional change centered on evolution; second, to provide examples of intentional behavioral and cultural change from the applied behavioral sciences, which are largely unknown to the basic scientific community.

All species have evolved mechanisms of phenotypic plasticity that enable them to respond adaptively to their environments. Some mechanisms of phenotypic plasticity count as evolutionary processes in their own right. The human capacity for symbolic thought provides an inheritance system with the same kind of combinatorial diversity as genetic recombination and antibody formation. Taking these propositions seriously allows an integration of major traditions within the basic behavioral sciences, such as behaviorism, social constructivism, social psychology, cognitive psychology, and evolutionary psychology, which are often isolated and even conceptualized as opposed to each other.

The applied behavioral sciences include well-validated examples of successfully managing behavioral and cultural change at scales ranging from individuals, to small groups, to large populations. However, these examples are largely unknown beyond their disciplinary boundaries, for lack of a unifying theoretical framework. Viewed from an evolutionary perspective, they are examples of managing evolved mechanisms of phenotypic plasticity, including open-ended processes of variation and selection.

Once the many branches of the basic and applied behavioral sciences become conceptually unified, we are closer to a science of intentional change than one might think.

1. Introduction

Change is the mantra of modern life. We embrace change as a virtue but are desperate to escape from undesired changes that appear beyond our control. We crave positive change at all levels: individuals seeking to improve themselves, neighborhoods seeking a greater sense of community, nations attempting to function as corporate units, the multinational community attempting to manage the global economy and the environment.

Science should be an important agent of change, and it is; but it is responsible for as many unwanted changes as those we desire. Even the desired changes can be like wishes granted in folk tales, which end up regretted in retrospect. Despite some notable successes, some of which we highlight in this article, our ability to change our behavioral and cultural practices lags far behind our ability to manipulate the physical environment. No examples of scientifically guided social change can compare to putting a man on the moon.

In this article we ask what a science of positive behavioral and cultural change would look like and what steps might be required to achieve it. We begin with the basic suggestion that evolution must be at the center of any science of change. After all, evolution is the study of how organisms change in relation to their environments, not only by genetics but also by mechanisms of phenotypic plasticity that evolved by genetic evolution, including some that count as evolutionary processes in their own right (Calvin 1987; Jablonka & Lamb 2006; Richerson & Boyd 2005). A solid foundation in evolutionary theory can also help us understand why some changes we desire, which count as adaptations in the evolutionary sense of the word, can turn out to be bad for long-term human welfare. Left unmanaged, evolutionary processes often take us where we would prefer not to go. The only solution to this problem is to become wise managers of evolutionary processes (Wilson 2011c).

The first step – appreciating the central importance of evolution – reveals how many steps remain to achieve a mature science of behavioral and cultural change. The study of evolution in relation to human affairs has a long and tortuous history that led many to abandon and even oppose the enterprise altogether (Ehrenreich & McIntosh 1997; Sahlins 1976; Segerstrale 2001). Using evolution to inform public policy earned such a bad reputation that “social Darwinism” came to signify the justification of social inequality (Hofstadter 1959/1992; Leonard 2009). Evolution became a pariah concept to avoid as a conceptual foundation for the study of human behavior and culture for most of the 20th century. The implicit assumption was that evolution explained the rest of life, our physical bodies, and a few basic instincts such as the urge to eat and have sex, but had little to say about our rich behavioral and cultural diversity.

The reception to E. O. Wilson’s 1975 book Sociobiology provides an example of this intellectual apartheid. The purpose of Sociobiology was to show that a single science of social behavior could apply to all species, from microbes to insects to primates. It was celebrated as a triumph except for the final chapter on humans, which created a storm of controversy (Segerstrale 2001). Only during the late 1980s did terms such as evolutionary psychology and evolutionary anthropology enter the scientific language, signifying a renewed attempt to place the study of human behavior and culture on an evolutionary foundation.

As a result, an enormous amount of integration must occur before a science of human behavioral and cultural change can center on evolution. This integration needs to be a two-way street, involving not only contributions of evolutionary theory to the human-related disciplines but also the reverse. For example, core evolutionary theory needs to expand beyond genetics to include other inheritance systems, such as environmentally induced changes in gene expression (epigenetics), mechanisms of social learning found in many species, and the human capacity for symbolic thought that results in an almost unlimited variety of cognitive constructions, each motivating a suite of behaviors subject to selection (Jablonka & Lamb 2006; Penn et al. 2008).

We will argue that the first steps toward integration, represented by a configuration of ideas that most people associate with evolutionary psychology, was only the beginning and in some ways led in the wrong direction. In particular, the polarized distinction between evolutionary psychology and the standard social science model (Pinker 1997; 2002; Tooby & Cosmides 1992) was a wrong turn we must correct. A mature EP needs to include elements of the SSSM associated with major thinkers such as Emile Durkheim, B. F. Skinner, and Clifford Geertz. Only when we depolarize the distinction between EP and the SSSM can a science of change occur (Bolhuis et al. 2011; Buller 2005; Scher & Rauscher 2002; Wilson 2002b).

In section 2 of this article we will attempt to accomplish this depolarization to provide a broader evolutionary foundation for the human behavioral and social sciences. In section 3 we will review examples of scientifically based and validated programs that accomplish change on three scales: individuals, small groups, and large populations. We draw these examples from branches of the applied behavioral sciences that, like diamonds in the sand, have remained largely hidden from evolutionary science and the basic human behavioral sciences. The examples provide a much needed body of empirical information to balance evolutionary theorizing, which is frequently criticized for remaining at the speculative “just so” storytelling stage. Indeed, the randomized control trials and other high-quality real-world experiments described in section 2 can be regarded as a refined variation-and-selection process with faster and more accurate feedback on effectiveness than other mechanisms of cultural evolution. When viewed from an evolutionary perspective, they emerge as examples of wisely managing evolutionary processes to accomplish significant improvement in human well-being. We are closer to a science of intentional change than one might think.

2. Toward a basic science of change

The ability to change behavioral and cultural practices in practical terms can profit from a basic scientific understanding of behavioral and cultural change. The human behavioral sciences are currently in disarray on the subject of change. Every discipline has its own configuration of ideas that seldom relate to other disciplines or to modern evolutionary science. We will focus on a major dichotomy that all human-related disciplines must confront: On the one hand, human behavior and culture appear elaborately flexible. On the other, as with all species, the human brain is an elaborate product of genetic evolution. These two facts often appear in opposition to each other, as if evolution implies genetic determinism, which in turn implies an incapacity for change over short time intervals. Once this misformulation is accepted, then the capacity for short-term change becomes conceptualized as outside the orbit of evolutionary theory.

Although the tension between genetic innateness and the capacity for short-term change exists in all branches of the human behavioral sciences, we will focus on two major branches: the behaviorist tradition associated with B. F. Skinner and the configuration of ideas that arose in the late 1980s under the label evolutionary psychology (EP). These merit special attention because of the history of the behaviorist tradition in academic psychology, even before EP made the scene, and because EP came about in a way that seemed to exclude the standard social science model (SSSM) centered on behaviorism in psychology and so-called blank slate traditions in anthropology associated with figures such as Durkheim and Geertz (e.g., Pinker 1997; 2002; Tooby & Cosmides 1992). Reconciling the differences between the behaviorist tradition and EP can go a long way toward reconciling the apparent paradox of genetic innateness and the capacity for short-term change in all branches of the human behavioral sciences.

2.1. B. F. Skinner: Evolutionary psychologist

In the abstract of his influential article “Selection by Consequences,” Skinner (1981) framed his version of behaviorism in terms of evolution:
Selection by consequences is a causal mode found only in living things, or in machines made by living things. It was first recognized in natural selection, but it also accounts for the shaping and maintenance of the behavior of the individual and the evolution of cultures. In all three of these fields, it replaces explanations based on the causal modes of classical mechanics. The replacement is strongly resisted. Natural selection has now made its case, but similar delays in recognizing the role of selection in the other fields could deprive us of valuable help in solving the problems which confront us. (p. 501)
Although the term evolutionary psychology had not yet been coined, Skinner’s passage leaves no doubt that he regarded the open-ended capacity for behavioral and cultural change as both (1) a product of genetic evolution and (2) an evolutionary process in its own right. It is therefore ironic that when Tooby and Cosmides (1992) formulated their version of EP, they set it apart from the SSSM that included the Skinnerian tradition (see also Pinker 1997; 2002). 

Long before Tooby and Cosmides’s version of EP made the scene, the so-called cognitive revolution had largely displaced behaviorism in academic psychology. Cognitive theorists stressed that the enormous complexity of the mind needed direct study, in contrast to Skinner’s insistence on focusing on the functional relations of environment and behavior (Brewer 1974; Bruner 1973). The central metaphor of the cognitive revolution was that the mind is like a computer that we must understand in mechanistic detail to know how it works. However, those who study computers would never restrict themselves to input-output relationships: They would study the machinery and the software. Cognitive psychologists faulted behaviorists for not following the same path.

One of the seeds of the cognitive revolution, which took root in Tooby and Cosmides’s version of EP, was a challenge to what most perceived to be the extreme domain generality of behavioral approaches. An example is Martin Seligman’s (1970) influential article on the “generality of the laws of learning.” Seligman reviewed a body of evidence showing that the parameters of learning processes had to be viewed in light of the evolutionary preparedness of organisms to relate particular events. For example, taste aversion (Garcia et al. 1966) challenged the idea that immediacy per se is key in stimulus pairings in classical conditioning, as illness could follow by tens of hours and still induce aversion to ecologically sensible food-related cues. Seligman recognized that this kind of specialized learning could evolve by altering the parameters of classical conditioning, but his preferred interpretation was that general learning processes themselves were not useful: “[W]e have reason to suspect that the laws of learning discovered using lever pressing and salivation may not hold” (p. 417).

Even more important was the conclusion that no general process account was possible in the area of human language and cognition. Pointing to evidence that seemed to show that human language requires no elaborate training for its production, Seligman concluded, “instrumental and classical conditioning are not adequate for an analysis of language” (p. 414). What interests us in this context is how these concerns quickly led to abandoning the idea that general learning process accounts were possible. For example, in an influential chapter that helped launch the “cognitive revolution,” William Brewer (1974) concluded, “all the results of the traditional conditioning literature are due to the operation of higher mental processes, as assumed in cognitive theory, and … there is not and never has been any convincing evidence for unconscious, automatic mechanisms in the conditioning of adult human beings” (p. 27, italics added).

The concern over the limits of domain generality in cognitive psychology redoubled as EP arrived as a self-described discipline, including the influential volume The Adapted Mind: Evolutionary Psychology and the Generation of Culture (Barkow et al. 1992; see also Pinker 1997; 2002). The thrust of EP was that the mind is neither a blank slate nor a general-purpose computer. The mind is a collection of many special-purpose computers that evolved genetically to solve specific problems pertaining to survival and reproduction in ancestral environments. This claim became known as “massive modularity” (Buller 2005; Buller & Hardcastle 2000; Carruthers 2006; Fodor 1983; 2000).

Tooby and Cosmides’s (1992) chapter in The Adapted Mind, titled “The Psychological Foundations of Culture,” which did much to define the field of EP, described domain-general learning (the applicability of general cognitive processes, whether viewed behaviorally or cognitively) as nearly a theoretical impossibility. Too many environmental inputs can be processed in too many ways for a domain-general learning machine to work, whether designed by humans or by natural selection. The most intelligent machines humans have designed are highly task specific. Tax preparation software provides a good example: It requires exactly the right environmental input, which it processes in exactly the right way, to calculate one’s taxes accurately. It is impressively flexible at its specific task but utterly incapable of doing anything else. According to Tooby and Cosmides, natural selection is constrained just as human engineers are in creating complex machines or programming software, leaving massive modularity as the only theoretical possibility for the design of the mind.

In discussing cultural evolution, Tooby and Cosmides observed that behavioral differences among human populations do not necessarily signify the cultural transmission of learned information. Instead, they can reflect massively modular minds responding to different environmental cues without any learning or social transmission whatsoever. They called this instinctive response to the environment “evoked” culture, in contrast to the social transmission of learned information, or “transmitted” culture. They did not deny the existence of transmitted culture, but had little to say about it.

An article titled “Evolutionary Psychology: A Primer” (Cosmides & Tooby 1997) pares their vision to its bare essentials. The human mind is described as “a set of information processing machines that were designed by natural selection to solve adaptive problems faced by our hunter-gatherer ancestors.” Because our modern skull houses a Stone-Age mind, “the key to understanding how the modern mind works is to realize that its circuits were not designed to solve the day-to-day problems of a modern American – they were designed to solve the day-today problems of our hunter-gatherer ancestors.” Evolutionary psychology is described as “relentlessly past-oriented” – meaning our genetic past, not our cultural or individual past.

In this fashion, the concept of elaborate innateness that became associated with EP sat in opposition to the open-ended capacity for change that became associated with what Tooby and Cosmides branded the SSSM. In our opinion, this is a profound mistake needing correction to achieve an integrated science of change.
Read the whole article.

Nepal: In the Mountain's Shadow

This is both a sad and beautiful documentary largely about the efforts of one man, Visma Raj Paudel, to provide refuge for the many orphans left by the on-going conflicts between the Maoist revolutionaries and the more progressive and educated elite.

Nepal: In the Mountain's Shadow

2009 | 48 minutes

Nepal is home to the Himalayan Mountains. For thousands of years people have traveled to the Himalayas seeking spiritual enlightenment, proclaiming man could be freed of all sin by merely gazing upon their peaks. This tradition is continued today by pilgrims who journey to Nepal across the globe to glimpse upon its natural beauty and explore its ancient history. But there's another side to Nepal. A side few travelers ever witness. Poverty-stricken slums and villages have become a common sight across the landscape.

The main role of the government of any country is to address the problems of the people and to find better solutions, but that's not happening in Nepal. The poor become poorer and poorer, and the rich become richer and richer, and there's a big gap... while the government isn't providing even the basic infrastructure to these poor people. Without further education most Nepalese are condemned to a life of manual labor earning an average income of $300 per year.

For decades Nepal's leaders have struggled to provide for their people. The situation escalated in the mid 1990s when civil war broke out across the country. And what started as a movement toward democracy ended in catastrophe. Over 12,000 people were killed in the conflict.

Many years later Visma Raj Paudel opened the first of the many projects to come... a children's orphanage. Refusing to give up on his dreams, Visma began seeking alternative means to help the people and children of Nepal. Today the orphanage is home to more than 70 children. It's no wonder that when Visma first began selecting children for the orphanage one of the first places he went back to was his childhood village.

To date the orphanage has received over 1,000 applications from communities open to provide children with a better life. Unable to adopt them all, Visma started the scholarship program. He's supporting only 200 children under his different projects, but it's very true that are so many children on the road and in the villages who are abandoned, not going to school, and forced to work as child servants.

Watch the full documentary now

Friday, May 16, 2014

Modulation of Adult Hippocampal Neurogenesis by Early-Life Environmental Challenges Triggering Immune Activation

By way of introduction to this interesting article, I am going to basically reproduce much of the abstract, since it summarizes the article better than can I.
Research has shown that factors such as maternal stress and nutrition as well as maternal infections can activate the immune system in the infant. A rising number of research studies have shown that activation of the immune system in early life can augment the risk of some psychiatric disorders in adulthood, such as schizophrenia and depression. The mechanisms of such a developmental programming effect are unknown; however some preliminary evidence is emerging in the literature, which suggests that adult hippocampal neurogenesis may be involved. A growing number of studies have shown that pre- and postnatal exposure to an inflammatory stimulus can modulate the number of proliferating and differentiating neural progenitors in the adult hippocampus, and this can have an effect on behaviours of relevance to psychiatric disorders. This review provides a summary of these studies and highlights the evidence supporting a neurogenic hypothesis of immune developmental programming.
This research is a crucial piece in the puzzle of understanding how early pre-, peri-, and post-natal experience, especially stress (which causes inflammation), can impact adult brain function. 

Full Citation:
Musaelyan, K, Egeland, M, Fernandes, C, Pariante, CM, Zunszain, PA, and Thuret, S. (2014, May 7). Modulation of Adult Hippocampal Neurogenesis by Early-Life Environmental Challenges Triggering Immune Activation. Neural Plasticity; Article ID 194396, 10 pages. doi:

Modulation of Adult Hippocampal Neurogenesis by Early-Life Environmental Challenges Triggering Immune Activation

Ksenia Musaelyan [1,2,3], Martin Egeland [1,2], Cathy Fernandes [3], Carmine M. Pariante [1,2], Patricia A. Zunszain [1,2], and Sandrine Thuret [1]
1. Centre for the Cellular Basis of Behaviour, Institute of Psychiatry, King’s College London, The James Black Centre, 125 Coldharbour Lane, London SE5 9NU, UK
2. Department of Psychological Medicine, Institute of Psychiatry, King’s College London, The James Black Centre, 125 Coldharbour Lane, London SE5 9NU, UK
3. MRC Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, King’s College London, 16 De Crespigny Park, London SE5 8AF, UK
Copyright © 2014 Ksenia Musaelyan et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 


The immune system plays an important role in the communication between the human body and the environment, in early development as well as in adulthood. Per se, research has shown that factors such as maternal stress and nutrition as well as maternal infections can activate the immune system in the infant. A rising number of research studies have shown that activation of the immune system in early life can augment the risk of some psychiatric disorders in adulthood, such as schizophrenia and depression. The mechanisms of such a developmental programming effect are unknown; however some preliminary evidence is emerging in the literature, which suggests that adult hippocampal neurogenesis may be involved. A growing number of studies have shown that pre- and postnatal exposure to an inflammatory stimulus can modulate the number of proliferating and differentiating neural progenitors in the adult hippocampus, and this can have an effect on behaviours of relevance to psychiatric disorders. This review provides a summary of these studies and highlights the evidence supporting a neurogenic hypothesis of immune developmental programming.

1. Introduction

Adult hippocampal neurogenesis is a complex process which includes continuous renewal of the hippocampal stem cell pool and generation of new neurons in the dentate gyrus (DG) of the hippocampus during adult life [1]. As neurogenesis is considered to be an important contributor to the neuroplastic ability of the hippocampus, it is suggested that it can be influenced by multiple environmental factors. The immune system plays a key role in the communication between the body and the environment [2]. Therefore investigating the relationship between the immune system and hippocampal neurogenesis can provide an important insight into the impact of environmental factors on a neurogenic niche, a region where neurogenesis occurs in a regulatory microenvironment. Recent research indicates that adult hippocampal neurogenesis is indeed extensively affected by elements of the immune response such as activated microglia and cytokine release. Immune activation has been shown to play an important role in many neuropsychiatric diseases such as Alzheimer’s disease, schizophrenia, autism, and depression [3]. Importantly, a growing number of studies report that immune activation in utero can augment the risk of these disorders in adulthood. Accordingly, maternal infections have been associated with increased risk of schizophrenia and autism [4, 5]. In depression, early-life factors which are known to increase the risk of the disorder in adulthood, such as maternal anxiety and perinatal depression as well as traumatic postnatal life events, are associated with changes in the immune system [3, 6, 7]. This evidence suggests that environmental factors in early-life predispose individuals to the development of psychopathological conditions in adulthood, potentially via the activation of the immune system. The mechanism of such developmental programming is unknown; however some preliminary evidence is emerging in the literature which suggests that adult hippocampal neurogenesis may be involved. The neurogenic hypothesis of immune developmental programming is still in its early days and this review will summarize the emerging supporting evidence available to date.

2. The Effect of Inflammation on Hippocampal Neurogenesis in Adult Life

The relationship between inflammatory factors and adult hippocampal neurogenesis has been a focus of multiple studies during recent years. Most of the knowledge in this area comes from studies which used lipopolysaccharide (LPS) or cytokine administration to model the inflammatory state. LPS is a bacterial endotoxin which is commonly used to reproduce an immune response where its systemic administration induces a behavioural phenotype known as “sickness behavior,” through an increase of the levels of proinflammatory cytokines in peripheral blood as well as in the brain [8]. The central effect of LPS can be attributed to the activation of microglia, which is known to release cytokines in the activated state. In 2003, Ekdahl and coworkers [9] demonstrated that direct intracortical administration of LPS dramatically reduced the survival of newly generated neurons in the adult hippocampus, while Monje and coworkers found the same effect of peripheral LPS administration [10]. Other studies supported the role of individual cytokines in this process, demonstrating that overexpression of interleukin 6 (IL-6) and interleukin 1 beta (IL-1β) in the brain as well as systemic administration of tumor necrosis factor alpha (TNF-α) can reduce cell proliferation and neuronal differentiation in the adult hippocampus, stimulating more progenitor cells to differentiate into astrocytes [11–13]. An in vitro study which employed human hippocampal stem cell line confirmed that this effect also occurs in human hippocampal cells [14]. Moreover it appears that not only the production, but also the function of new neurons is affected by inflammation. Accordingly, Belarbi and coworkers showed that chronic exposure to LPS decreased the recruitment of new neurons into hippocampal networks following training in a spatial exploration task [15]. This further suggests that neurogenesis might be responsible for the cognitive deficits seen in infectious diseases accompanied by inflammation. Currently, microglial activation and cytokine release are also suggested to be an underlying mechanism of neurogenic decline in noninfectious conditions such as Alzheimer’s disease and normal aging [16].

3. Programming Effects of Early-Life Immune Activation

3.1. Early-Life Immune Activation in Humans: Consequences for Adult Behaviour and Psychopathology
Studies have shown that inflammation plays a specific neuroregulatory role not only in adult life and aging, but also in early development. Thus clinical evidence suggests that prenatal exposure to maternal infections predisposes an individual to the development of neuropsychiatric diseases such as schizophrenia and autism [5]. However the causality of immune activation extends beyond infectious conditions. It has been shown that other environmental factors apart from exposure to infectious agents can activate the immune system in mother and child. Nutritional status of mother-infant dyad can have a profound effect on the developing immune system. While maternal nutrition during pregnancy insures sufficient availability and transfer of immune factors to the foetus and can influence the development of the immune system, postnatal nutrition has an additional function of introducing the infant’s immune system to food antigens and subsequent development of optimal immune tolerance of the gut. Impaired immune tolerance can lead to gut inflammatory disease and food allergies, conditions which lead to chronic activation of the immune system [17]. Moreover, food is one of the most important sources of environmental contaminants, toxic compounds created by industrial activity. Exposure to these xenobiotics can lead to developmental immunotoxicity (reviewed in [6]). Importantly, psychosocial factors can also affect immune system in the early-life. Thus maternal stress and anxiety have been shown to affect immune communication between mother and foetus which in turn affects the postnatal immunological status of the child. As such, maternal anxiety has been associated with reduced adaptive immunity in infants [18], while prenatal exposure to stress has been linked to increased risk of infectious disease-related hospitalisation in childhood [19]. Similarly for postnatal stress, Danese and colleagues described a positive association between childhood maltreatment and an increase in the inflammatory marker C-reactive protein in adulthood [7].

3.2. Early-Life Immune Activation in Animal Models: Consequences for Adult Immune and Stress Response Regulation
While human data in this field are still quite limited, animal studies provide abundant evidence demonstrating the link between perinatal immune activation and consequences for the immune system and stress susceptibility in adulthood. These studies, in line with the inflammation and neurogenesis studies described above, also use LPS to activate the immune system; however E. coli infection was employed by some research groups to stimulate immune response [20–22]. Most of the studies show that immune stimulation over the first few days of life leads to changes in the cytokine production in the brain in response to LPS exposure in adulthood. Thus Bilbo and colleagues showed that neonatal E. coli infection at PND 4 led to an increase in IL-1β protein shortly after (1.5 hrs) adult LPS exposure. Interestingly this effect was specific for the hippocampus and parietal cortex [21], while hypothalamic IL-1β protein content was decreased in the same conditions [22]. These changes also correlated with hippocampal function, as neonatally exposed rats displayed memory impairment in the contextual memory task following LPS injection in adulthood [22]. Furthermore, neonatally infected rats had persistently elevated levels of microglial activation markers such as major histocompatibility complex II (MHCII) and Iba-1 in the hippocampus in adulthood [20, 22, 23]. This increase may suggest that hippocampal microglia remained in a “primed” state during adulthood, characterised by an exaggerated proinflammatory response when exposed to the immune challenge [24]. Interestingly, the direction of observed change in cytokine content might be time- and dose-specific. This is exemplified in a study by Kohman et al. who observed an opposing decrease in hippocampal IL-1β gene expression looking specifically at expression in 4 hours (rather than 1.5 hours) after LPS injection in an otherwise similar study design [25]. While this decrease can be a compensatory response to an elevation of the protein content observed at 1.5 hrs after the injection [21], it is also possible that the difference in the direction of the response is due to the dose difference between the studies, as the dose of LPS used by Kohman et al. to challenge immune system in adulthood is 10 times higher than that described in the previous studies [20–22]. Nonetheless, these studies indicate that neonatal infection can modulate adult immune response in the hippocampus. Interestingly, none of these studies found change in peripheral cytokine levels in neonatally exposed animals. However, a study which exposed rat pups to LPS at a later time point, PND 14, described an attenuated elevation of blood cytokines in response to LPS challenge in adulthood [26]. Such discrepancy demonstrates that the timing of postnatal exposure can significantly influence the type of response observed in adulthood. Indeed it has been shown that some of the proteins involved in immune response to antigens such as LPS-binding protein (LBP) are not expressed during the first week of life [27]. Interestingly LBP expression peaks at PND 14, which might explain the more robust effects observed in Ellis et al.’s study described previously.

Early-life inflammatory challenges have also been shown to cause behavioural changes in adulthood. This is exemplified by how the stress response in adulthood has been shown to be modulated by early-life LPS exposure with many studies showing increased fear and anxiety-like behaviour and decreased exploratory behaviour in relevant tasks such as elevated plus maze and open field/hide box in response to restraint stress [28–30]. The underlying mechanism of this outcome is suggested to be the programming effects of early-life LPS exposure on hypothalamo-pituitary-adrenal (HPA) axis regulation. This notion is supported by observed changes in endocrine factors on each level of the HPA axis. More specifically, increased baseline corticotropin-releasing hormone (CRH) expression in the paraventricular nucleus of the hypothalamus [31], increased plasma concentrations of adrenocorticotrophic hormone (ACTH), increased and/or prolonged adrenal corticosterone release in response to stress in neonatally LPS-exposed animals [31–34], and decreased glucocorticoid receptor (GR) density in the hypothalamus, hippocampus, and frontal cortex of adult animals, where GR is thought to play a key role in negative feedback regulation of the HPA axis [31].

Further aspects of behaviour such as reproductive function and memory have also been shown to be affected. Thus Wu et al. showed that female rats exposed to LPS in early-life displayed a significant delay in puberty and disruption of oestrous cycle which persisted into adulthood [35]. Moreover it has been demonstrated that neonatal LPS exposure leads to a decrease in inflammatory stress resilience of the female reproductive system by long-term sensitization of gonadotropin releasing hormone (GnRH) regulator, a central controller of reproduction. This effect was reflected by increased suppression of luteinising hormone frequency in response to adult LPS exposure [36]. In male animals, early-life LPS exposure also has been shown to affect reproductive function. Accordingly, Walker et al. demonstrated that neonatally exposed male animals show reduced sexual activity following stress, indicated by reduced amount of mounts and ejaculation when presented with a female [37].

Learning- and memory-related behaviours have also been reported to be affected by prenatal or neonatal inflammatory agent exposure. Specifically, Kohman et al. showed that neonatally exposed rats show associative learning and memory impairment in active avoidance conditioning task [25], while a similar effect was observed in this task for adult offspring of mice subjected to poly I : C injections during pregnancy [38].

Thus, research demonstrates that early-life exposure to inflammatory stimuli modulates multiple aspects of development from the immune system to the stress response and, furthermore, that some of these effects are specific for the hippocampus. Importantly both immune system activation and chronic stress are implicated in psychopathological conditions such as anxiety and depression. Therefore, investigating the mechanisms through which early-life immune activation exerts such a long-lasting effect on these systems might shed light on the means by which early-life environmental insults predispose individuals to the development of psychopathology in adult life.

As some of the changes found to be affected by inflammation early in development were specific for the hippocampus, it follows that the hippocampus would be a first target in search of the underlying mechanism. Indeed, the hippocampus is an interesting brain area as it has been recently confirmed that it is the only area of the human brain where neurogenesis occurs throughout adult life [39]. Hippocampal neurogenesis has been extensively implicated in mood disorders and in the mechanism of antidepressant action [40, 41]. Therefore it appears that modulation of neurogenesis by early-life immune activation might play a role in the long lasting effects of the perinatal inflammatory insult. However, the study of early-life inflammatory modulation of adult hippocampal neurogenesis has only recently attracted researchers’ attention. The next section will describe the emerging studies published to date which introduce this new important field of neurogenesis research and suggest directions for future investigation.

4. Perinatal Immune Activation and Adult Hippocampal Neurogenesis

4.1. Methodological Considerations in the Study of the Effect of Perinatal Immune Activation on Adult Hippocampal Neurogenesis
Studies investigating the effect of early-life inflammation on neurogenesis have employed similar models to activate the immune response as those dedicated to the effects of inflammatory challenge on immune response or behaviour in adulthood. LPS injection is a common way to activate immune response, as LPS is recognised by toll-like receptor 4 (TLR4) on antigen-presenting cells and effectively models bacterial infection. In the case of prenatal exposure, maternal antigen-presenting cells may interact with the placenta, capable of producing its own cytokines, which then enter foetal bloodstream [24]. A viral mimetic poly I : C also has been used in some of the studies. Poly I : C stimulates TLR3 receptor which recognises viral antigens. As it has been shown that maternal influenza increases the risk of schizophrenia in the offspring [4], poly I : C also appears to be a valid model to study long-term effects of perinatal inflammation.

Timing of the inflammatory exposure is another crucial point in the study design, as some reports suggest that different phases of pregnancy represent unique vulnerability windows to environmental inputs [42]. Thus the rodent prenatal period can be approximately correlated with human pregnancy as gestation days 1–9 (GD 1–9) being representative of first trimester of human pregnancy, GD 10–19 as a second trimester, and GD 19-20 to postnatal day (PND) 7 as corresponding to the third trimester of pregnancy [43]. It is important to note however that an exact comparison is not possible as the course of rodent and human pregnancy and brain development differ significantly; therefore some discrepancies exist between research groups in the interpretation of gestational timing.

The timing of the assessment of hippocampal neurogenesis is also an important factor to consider in the study design. Thus neurogenesis assessed before PND 21, when offspring is usually weaned from the mother, will represent the state of juvenile development, the period between PND 21 and 50, a period when rodents reach reproductive maturity [37], as adolescence and early adulthood, while PND 60 onwards is generally agreed to be representative of adulthood [44].

Studies also differ in which aspects of the neurogenesis were assessed and the methodological approaches used to characterise them. For example, cell proliferation assay, usually assessed using bromodeoxyuridine (BrdU) incorporation or KI67 immunohistochemistry, shows the rate of proliferation of the neural progenitors within the dentate gyrus. This is an important marker of the renewal of the pool of neural progenitors, which is thought to be crucial to maintain neurogenic capacity of the hippocampus throughout life [45]. However, this assay does not provide any information on the subsequent fate of the new cells. The survival rate (usually measured by the proportion of BrdU+ neurons which have survived in the hippocampus for 28 days after BrdU incorporation) shows whether proliferating cells have survived for a period of time which is long enough to acquire a mature phenotype. Studies vary in the timing of cell survival assessed and therefore interstudy data comparisons should be done with caution. A differentiation assay with doublecortin (DCX) can show the amount of neuroblasts present in the hippocampus. This parameter represents the number of newly born neurons which have a potential to be incorporated into hippocampal networks, an important measure of the neuroplastic potential of the hippocampus owing to neurogenesis. Finally neural progenitor fate mapping shows the proportion of newly born cells which became neurons or astrocytes. It provides an insight into the neurogenic versus astrogenic balance of the differentiation in the dentate gyrus, which is thought to change depending on regulatory signals received by the hippocampus from the peripheral environment.

In summary, it is important to be cautious when comparing studies using different methodological approaches in the field of early-life effects on hippocampal neurogenesis, as even small variations in methods used can produce substantial differences between the models. However such variety also allows us to study the topic from different developmental and cellular points of view, which can ultimately provide a more comprehensive understanding of observed effects.

4.2. The Effect of Perinatal Immune Activation on Adult Hippocampal Neurogenesis: Research Conducted to Date
Studies investigating the early-life immune modulation of adult hippocampal neurogenesis available to date are summarised in Table 1. One of the first studies in the field was done by Meyer and colleagues [42]. In this study pregnant C57BL/6 mice were injected with poly I : C at GD 9 and GD 17 which, according to the authors, corresponded to the early-to-mid and mid-to-late periods of pregnancy. Immunohistochemistry using DCX marker showed a decrease in the number of young neurons in the dentate gyrus of prenatally exposed offspring at PND 24, which corresponds to the early adolescent period. This effect was independent of the timing of the prenatal treatment. However, behavioural effects observed in adulthood of the offspring showed a different picture. Behavioural tests done on the mice at the age of 14–16 weeks demonstrated that mice exposed to the maternal poly I : C response on GD 9 displayed increased anxiety-like behaviour in the open field test, while mice exposed to GD 17 expressed behaviours indicative of perseverations in the discrimination reversal learning task. These data are of interest as they suggest that immune activation at different stages of prenatal development can result in differential outcomes for behavioural abnormalities in adult life. It is interesting to speculate that behavioural differences in adulthood could be accompanied by the differences in neurogenesis which perhaps were not yet apparent in the adolescent period. Even though the measured parameters of neurogenesis in this study were limited, additional factors assessed in the hippocampus of these mice provided an insight into mechanisms which might cause the observed reduction in the number of young neurons. Indeed, DCX decrease was accompanied by the decrease in reelin, a protein which is known to regulate neurogenesis [46]. This evidence led the authors to suggest that the observed reduction could be due to increased apoptosis of progenitor cells. However, analysis of the apoptotic marker caspase-3 demonstrated that increased caspase-3-dependant apoptosis only took place in the hippocampi of GD 17 exposed animals, yet another indication of the time-dependant variation in the response to an immune challenge. The authors suggested that such variation was initially due to the differential foetal immune response to maternal inflammation. Specifically, they showed that while poly I : C injections resulted in an increase in cytokine levels in the brains of both GD 9 and GD 17 exposed foetuses, only GD 17 group had an accompanying increase in cytokine gene expression, suggesting that cytokine increase was due to endogenous production by the foetal brain in this group only.
Table 1: A summary of the studies investigating the effect of perinatal immune activation on postnatal hippocampal neurogenesis. [View table]
A similar study design was later employed by Cui and colleagues [47]. This group also explored the consequences of maternal immune stimulation at different time points in pregnancy; however in their experimental design hippocampal neurogenesis was the primary focus of the study. More specifically, pregnant rats were peripherally administered LPS at a dose of 0.1 mg/kg on GD 15/16 (midgestation according to the study) or a dose of 0.05 mg/kg on GD 18/19 (late gestation). The dose difference was based on the increased sensitivity and mortality of near-term pregnant rats to LPS exposure. The results of BrdU incorporation assay showed that, by the end of the early perinatal period (the time between LPS injection and PND 14), the hippocampus of the GD 18/19 exposed offspring contained less cells which had undergone division, a decrease which might be due to decreased proliferation and/or survival of dividing cells. A similar trend was observed in the GD 15/16 group, but it did not reach significance. Interestingly, this situation was reversed when cell proliferation was assessed at PND 14: midgestation exposed animals had a significant decrease in hippocampal cell proliferation, while this trend did not reach significance in the late pregnancy exposed group. Furthermore, the number of cells which had undergone proliferation was significantly decreased in both conditions 4 weeks after BrdU incorporation, with the effect being more pronounced in the late gestation group. While this measure incorporates the effect of both cell proliferation and survival rates, this effect being stronger 4 weeks rather than 2 hours after BrdU incorporation suggests that indeed changes in cell survival rates might cause observed differences in the number of BrdU positive cells. Cell proliferation and survival in adult offspring (PND 60) were not affected; however the authors argued that even the deficit in proliferation and survival of neurons only in early development might have long term consequences, as some of these neurons could have been destined to survive and function for the lifetime of the individual. Finally, there was no change in neuronal differentiation in any of the described conditions.

These results were later challenged by another study which used prenatal LPS administration in pregnant rats. Graciarena and colleagues [44] employed a robust model of LPS administration which involved a higher dose of LPS (0.5 mg/kg) and repeated injections every other day between GD 14 and 20. Hippocampal neurogenesis was assessed only in adult offspring at a single time point of PND 60. Results showed that these animals had a consistent decrease in progenitor cell proliferation and neuronal differentiation into young and mature neurons arising from proliferating progenitors but the survival of proliferating cells was not affected. In addition, the authors also explored the role of microglia in the observed effects. They found that, in adult animals prenatally exposed to LPS, hippocampal microglia morphologically resembled stages II-III of activation. Interestingly the hippocampus was the only brain region where microglia displayed features of nonphagocytic activation, suggesting specific sensitivity of the hippocampus to LPS effects. The authors also explored cytokine expression in the hippocampus. At birth, prenatally exposed pups had increased IL-1β in the hippocampus (but not IL-6), as well as decreased anti-inflammatory cytokine transforming growth factor beta-1 (TGFβ1). IL-1β levels were restored by adulthood consistent with the morphological appearance of microglia (nonphagocytic activation). However the levels of TGFβ1 in the hippocampus remained reduced in adulthood of the exposed animals, which is of particular interest as TGFβ1 has been shown to possess proneurogenic properties [48]. Also, overexpression of TGFβ1 in the adult DG restored normal levels of neurogenesis in the LPS-exposed group, suggesting a causal role of TGFβ1 in LPS-related decrease in hippocampal neurogenesis. Furthermore the observed effects had a behavioural correlate as LPS-exposed animals displayed impaired recognition memory in the novel object recognition test, while TGFβ1 overexpression improved their performance in this memory task.

Moreover, in the follow-up study, the authors showed in an in vitro system that TGFβ1 exerts its proneurogenic effect through the canonical Smad 2/3 pathway [49]. This study also highlighted a unique ability of prenatal LPS exposure to cause long-term neurogenic changes. While adult LPS exposure produced an acute reduction in cell proliferation similar in magnitude to prenatal LPS effect, it failed to affect cell differentiation in the long term as did prenatal exposure. Consistent with the hypothesis of TGFβ1 involvement in these effects, TGFβ1 levels were not affected by adult LPS exposure [49].

The studies described above provide an initial insight into the mechanisms underlying long-term changes in hippocampal neurogenesis and some behavioural correlates for these changes. However, in order to establish a relevant hypothesis of immune developmental programming, a more direct link with clinically-relevant psychopathology is needed. Introducing such a link, a recent study by Lin and colleagues [50] employed a similar prenatal exposure to an LPS paradigm to look at a depression-like behaviour in adulthood. In this study pregnant rats were injected with LPS at a low dose of 0.066 mg/kg on GD 10.5. As in the studies described previously, the prenatally exposed adult offspring displayed changes in hippocampal neurogenesis. Neurogenic changes were assessed at two time points: during the juvenile period at PND 21 and in adult rats (PND 90). The changes included a reduction in cell proliferation at both time points and a decreased number of DCX positive neuroblasts. Consistent with previous studies, there was no difference in the survival of proliferating cells or in the number of proliferating cells becoming mature neurons. Furthermore, the authors addressed whether these changes would predispose adult offspring to depression-like behaviour. The results of behavioural testing showed that, in adulthood, these rats more readily displayed learned helplessness in the forced swim test, a pharmacologically established model of a depression-like behaviour. Importantly, the exposed offspring showed a trend towards higher susceptibility to anhedonia measured by the levels of sucrose consumption in the course of the chronic mild stress paradigm, a widely used model to induce depressive-like behaviour in rodents by environmental stress. Strengthening the clinical link even further, the authors also showed that observed behavioural and neurogenic changes were rescued by a chronic administration of a common antidepressant fluoxetine. Thus prenatally exposed fluoxetine treated adult animals did not show increased learned helplessness behaviour in the forced swim test and their levels of cell proliferation and young hippocampal neurons returned to control level. However it is important to note that fluoxetine also increased neurogenesis in control animals, suggesting that fluoxetine might not necessary rescue LPS-damaged pathways but might be acting through an alternative mechanism. Nonetheless authors also showed that fluoxetine reversed some cellular abnormalities in the hippocampus which were suggested to underlie the mechanism of neurogenic changes, specifically brain derived neurotrophic factor (BDNF) expression in the hippocampus and dendritic spine density of the dentate granule neurons. These results suggest that hippocampal neurogenesis is indeed involved in the ability of perinatal immune activation to cause behavioural abnormalities in adulthood and increase adulthood susceptibility to stress-related psychiatric disorders such as depression.

It is important to note that all above-mentioned studies used a model of prenatal exposure, limiting the interpretation of the findings to prenatal conditions. However, a recent study by Järlestedt et al. [51] extended the early-life inflammatory exposure model to the postnatal period. In this study C57BL/6 mice were injected with a relatively high dose of 1 mg/kg of LPS at PND 9. Measurement of hippocampal neurogenesis found that while there was no acute effect on hippocampal cell proliferation in these pups at PND 11, in the juvenile period (PND 41), these animals already displayed a decrease in cell proliferation and in the number of progenitors becoming mature neurons and astrocytes. In adulthood (PND 60) these animals had a decreased number of cells which had undergone proliferation since the LPS injection (BrdU positive), a measure which is affected by both proliferation and survival of neural progenitors. Interestingly, this study was the first to show regional differences in the effect of LPS exposure, as the decrease in the number of BrdU positive cells was specific for dorsal hippocampus. These data are in line with the emerging evidence of the distinct functions of neurogenesis in different regions of the hippocampus [52].

Studies described so far, although different in many aspects of study design, provide converging evidence of the neurogenic consequences of perinatal immune activation. A common trend emerging from the data suggests that perinatally exposed adult animals tend to have a compromised hippocampal neurogenesis with a decrease in cell proliferation and neuronal differentiation being described more often than changes in cell survival.

Interestingly, Jiang and colleagues recently published two studies which can potentially provide an insight into what precedes the decrease in hippocampal neurogenesis in adult animals [53, 54]. Although they used a familiar model of prenatal exposure of rats to inflammatory stimulus at GD 15, their results should be compared with caution to the studies described previously as an exposure to a bacterial suspension of E. coli was used to stimulate immune response. This model, although potentially more relevant for reproducing clinical infection, is specific to the pathogen employed; therefore the data obtained from these studies cannot be directly extended to the effect of other environmental factors such as stress or nutrition on the immune system.

In these studies the state of hippocampal neurogenesis was evaluated during the first month of pup life at PND 3, 7, 14, and 28. Interestingly the authors showed that cell proliferation was increased during the first two weeks of life with a peak at PND 7 but returned to control levels by PND 28. The authors also describe some evidence for the increase in BDNF and its receptor expression and protein levels which coincide with a proliferation increase, a finding which supports the role of BDNF in the effect of immune stimulation on neurogenesis suggested by Lin and Wang [50]. Data from this study provide preliminary evidence that early-life immune activation can initially lead to a compensatory increase in hippocampal neurogenesis in early development followed by a subsequent decrease in the levels of cell proliferation and neuronal maturation in adulthood.

5. Conclusion

To summarise the findings described previously, perinatal inflammatory stimulation has a potential to modulate hippocampal neurogenesis in adult life. More specifically, animal studies available to date show that exposure to immune activation in the pre- or postnatal period reduces the rate of cell proliferation and neuronal differentiation in the adult hippocampus. Studies described in this review also suggest some cellular mechanisms which can underlie this long-term modulation, such as nonphagocytic microglial activation, a decrease in anti-inflammatory cytokines, and neurotrophic factors in the hippocampus. However further research is needed to uncover comprehensive pathways in an effort to develop interventions which could reverse such detrimental modulation.

Importantly, an inflammatory stimulus in early-life can result not only from infectious agents, but also from other environmental factors, such as stress and nutrition. Therefore the suggested hypothesis of developmental modulation of adult hippocampal neurogenesis by early-life immune activation can represent a converging common pathway through which various factors of early-life environment modulate hippocampal function in adult life. Such a pathway is of particular importance to psychiatric research as adult hippocampal neurogenesis has been implicated in many psychiatric diseases which are known to be exacerbated by early-life adverse events. Importantly, current research suggests a number of life style aspects which positively influence hippocampal neurogenesis, among them are exercises such as running and dietary factors such as polyphenols and omega-3 unsaturated fatty acids [55, 56]. If the hypothesis presented in this review is confirmed, this could suggest a beneficial effect of the therapeutic interventions affecting hippocampal neurogenesis for the promotion of mental health in individuals who experienced early-life adversities.
Conflict of Interests

All the authors have received research funding from companies interested in developing anti-inflammatory strategies for the treatment of depression, such as Janssen Pharmaceutica, but the data reviewed in this paper are unrelated to this funding.


The authors of this paper receive research support from the following organizations: Janssen Pharmaceutica Studentship (KM), Marie Curie Actions Fellowship (ME), the UK Medical Research Council (MRC) (MR/J002739/1) (CMP), the South London and Maudsley NHS Trust and King’s College Hospital NIHR Biomedical Research Centre for Mental Health (CMP, PZ), MRC (MR/K022377/1), Eli Lilly and Company Ltd, Mount Sinai School of Medicine (0285-3965-4609), Autism Speaks (8132) (CF), The Psychiatry Research Trust, The Welton Foundation, and the MRC (ST).


  1. D. T. Balu and I. Lucki, “Adult hippocampal neurogenesis: regulation, functional implications, and contribution to disease pathology,” Neuroscience and Biobehavioral Reviews, vol. 33, no. 3, pp. 232–252, 2009.
  2. G. A. Rook, “Regulation of the immune system by biodiversity from the natural environment: an ecosystem service essential to health,” Proceedings of the National Academy of Sciences of the United States of America, vol. 110, no. 46, pp. 18360–18367, 2013.
  3. T. G. O’Connor, J. A. Moynihan, and M. T. Caserta, “Annual research review: the neuroinflammation hypothesis for stress and psychopathology in children—developmental psychoneuroimmunology,” Journal of Child Psychology and Psychiatry, 2013.
  4. A. S. Brown, “Prenatal infection as a risk factor for schizophrenia,” Schizophrenia Bulletin, vol. 32, no. 2, pp. 200–202, 2006.
  5. H. Ó. Atladóttir, P. Thorsen, L. Østergaard et al., “Maternal infection requiring hospitalization during pregnancy and autism spectrum disorders,” Journal of Autism and Developmental Disorders, vol. 40, no. 12, pp. 1423–1430, 2010.
  6. A. H. Marques, T. G. O’Connor, C. Roth, E. Susser, and A. L. Bjørke-Monsen, “The influence of maternal prenatal and early childhood nutrition and maternal prenatal stress on offspring immune system development and neurodevelopmental disorders,” Frontiers in Neuroscience, vol. 7, article 120, 2013.
  7. A. Danese, T. E. Moffitt, C. M. Pariante, A. Ambler, R. Poulton, and A. Caspi, “Elevated inflammation levels in depressed adults with a history of childhood maltreatment,” Archives of General Psychiatry, vol. 65, no. 4, pp. 409–415, 2008.
  8. S. Biesmans, T. F. Meert, J. A. Bouwknecht et al., “Systemic immune activation leads to neuroinflammation and sickness behavior in mice,” Mediators of Inflammation, vol. 2013, Article ID 271359, 14 pages, 2013.
  9. C. T. Ekdahl, J.-H. Claasen, S. Bonde, Z. Kokaia, and O. Lindvall, “Inflammation is detrimental for neurogenesis in adult brain,” Proceedings of the National Academy of Sciences of the United States of America, vol. 100, no. 23, pp. 13632–13637, 2003.
  10. M. L. Monje, H. Toda, and T. D. Palmer, “Inflammatory blockade restores adult hippocampal neurogenesis,” Science, vol. 302, no. 5651, pp. 1760–1765, 2003.
  11. J. A. Seguin, J. Brennan, E. Mangano, and S. Hayley, “Proinflammatory cytokines differentially influence adult hippocampal cell proliferation depending upon the route and chronicity of administration,” Neuropsychiatric Disease and Treatment, vol. 5, no. 1, pp. 5–14, 2009.
  12. L. Valliéres, I. L. Campbell, F. H. Gage, and P. E. Sawchenko, “Reduced hippocampal neurogenesis in adult transgenic mice with chronic astrocytic production of interleukin-6,” Journal of Neuroscience, vol. 22, no. 2, pp. 486–492, 2002.
  13. M. D. Wu, A. M. Hein, M. J. Moravan, S. S. Shaftel, J. A. Olschowka, and M. K. O'Banion, “Adult murine hippocampal neurogenesis is inhibited by sustained IL-1β and not rescued by voluntary running,” Brain, Behavior, and Immunity, vol. 26, no. 2, pp. 292–300, 2012.
  14. P. A. Zunszain, C. Anacker, A. Cattaneo et al., “Interleukin-1β: a new regulator of the kynurenine pathway affecting human hippocampal neurogenesis,” Neuropsychopharmacology, vol. 37, no. 4, pp. 939–949, 2012.
  15. K. Belarbi, C. Arellano, R. Ferguson, T. Jopson, and S. Rosi, “Chronic neuroinflammation impacts the recruitment of adult-born neurons into behaviorally relevant hippocampal networks,” Brain, Behavior, and Immunity, vol. 26, no. 1, pp. 18–23, 2012.
  16. R. A. Kohman and J. S. Rhodes, “Neurogenesis, inflammation and behavior,” Brain, Behavior, and Immunity, vol. 27, pp. 22–32, 2013.
  17. V. Verhasselt, “Oral tolerance in neonates: from basics to potential prevention of allergic disease,” Mucosal Immunology, vol. 3, no. 4, pp. 326–333, 2010.
  18. T. G. O’Connor, M. A. Winter, J. Hunn et al., “Prenatal maternal anxiety predicts reduced adaptive immunity in infants,” Brain, Behavior, and Immunity, vol. 32, pp. 21–28, 2013.
  19. N. M. Nielsen, A. V. Hansen, J. Simonsen, and A. Hviid, “Prenatal stress and risk of infectious diseases in offspring,” The American Journal of Epidemiology, vol. 173, no. 9, pp. 990–997, 2011.
  20. S. D. Bilbo, N. J. Newsum, D. B. Sprunger, L. R. Watkins, J. W. Rudy, and S. F. Maier, “Differential effects of neonatal handling on early life infection-induced alterations in cognition in adulthood,” Brain, Behavior, and Immunity, vol. 21, no. 3, pp. 332–342, 2007.
  21. S. D. Bilbo, L. H. Levkoff, J. H. Mahoney, L. R. Watkins, J. W. Rudy, and S. F. Maier, “Neonatal infection induces memory impairments following an immune challenge in adulthood,” Behavioral Neuroscience, vol. 119, no. 1, pp. 293–301, 2005.
  22. S. D. Bilbo, J. C. Biedenkapp, A. Der-Avakian, L. R. Watkins, J. W. Rudy, and S. F. Maier, “Neonatal infection-induced memory impairment after lipopolysaccharide in adulthood is prevented via caspase-1 inhibition,” Journal of Neuroscience, vol. 25, no. 35, pp. 8000–8009, 2005.
  23. L. Sominsky, A. K. Walker, L. K. Ong, R. J. Tynan, F. R. Walker, and D. M. Hodgson, “Increased microglial activation in the rat brain following neonatal exposure to a bacterial mimetic,” Behavioural Brain Research, vol. 226, no. 1, pp. 351–356, 2012.
  24. S. D. Bilbo and J. M. Schwarz, “The immune system and developmental programming of brain and behavior,” Frontiers in Neuroendocrinology, vol. 33, no. 3, pp. 267–286, 2012.
  25. R. A. Kohman, A. J. Tarr, N. L. Sparkman, T. M. H. Bogale, and G. W. Boehm, “Neonatal endotoxin exposure impairs avoidance learning and attenuates endotoxin-induced sickness behavior and central IL-1β gene transcription in adulthood,” Behavioural Brain Research, vol. 194, no. 1, pp. 25–31, 2008.
  26. S. Ellis, A. Mouihate, and Q. J. Pittman, “Early life immune challenge alters innate immune responses to lipopolysaccharide: implications for host defense as adults,” FASEB Journal, vol. 19, no. 11, pp. 1519–1521, 2005.
  27. L. Wei, A. Simen, S. Mane, and A. Kaffman, “Early life stress inhibits expression of a novel innate immune pathway in the developing hippocampus,” Neuropsychopharmacology, vol. 37, no. 2, pp. 567–580, 2012.
  28. G. Hava, L. Vered, M. Yael, H. Mordechai, and H. Mahoud, “Alterations in behavior in adult offspring mice following maternal inflammation during pregnancy,” Developmental Psychobiology, vol. 48, no. 2, pp. 162–168, 2006.
  29. A. K. Walker, T. Nakamura, R. J. Byrne et al., “Neonatal lipopolysaccharide and adult stress exposure predisposes rats to anxiety-like behaviour and blunted corticosterone responses: implications for the double-hit hypothesis,” Psychoneuroendocrinology, vol. 34, no. 10, pp. 1515–1525, 2009.
  30. F. R. Walker, J. March, and D. M. Hodgson, “Endotoxin exposure in early life alters the development of anxiety-like behaviour in the Fischer 344 rat,” Behavioural Brain Research, vol. 154, no. 1, pp. 63–69, 2004.
  31. N. Shanks, S. Larocque, and M. J. Meaney, “Neonatal endotoxin exposure alters the development of the hypothalamic- pituitary-adrenal axis: early illness and later responsivity to stress,” Journal of Neuroscience, vol. 15, no. 1, pp. 376–384, 1995.
  32. F. R. Walker, B. Knott, and D. M. Hodgson, “Neonatal endotoxin exposure modifies the acoustic startle response and circulating levels of corticosterone in the adult rat but only following acute stress,” Journal of Psychiatric Research, vol. 42, no. 13, pp. 1094–1103, 2008.
  33. M. H. Doosti, A. Bakhtiari, P. Zare et al., “Impacts of early intervention with fluoxetine following early neonatal immune activation on depression-like behaviors and body weight in mice,” Progress in Neuro-Psychopharmacology & Biological Psychiatry, vol. 43, pp. 55–65, 2013.
  34. A. K. Walker, T. Nakamura, and D. M. Hodgson, “Neonatal lipopolysaccharide exposure alters central cytokine responses to stress in adulthood in Wistar rats,” Stress, vol. 13, no. 6, pp. 506–515, 2010.
  35. X.-Q. Wu, X.-F. Li, B. Ye et al., “Neonatal programming by immunological challenge: effects on ovarian function in the adult rat,” Reproduction, vol. 141, no. 2, pp. 241–248, 2011.
  36. X. F. Li, J. S. Kinsey-Jones, A. M. I. Knox et al., “Neonatal lipopolysaccharide exposure exacerbates stress-induced suppression of luteinizing hormone pulse frequency in adulthood,” Endocrinology, vol. 148, no. 12, pp. 5984–5990, 2007.
  37. A. K. Walker, S. A. Hiles, L. Sominsky, E. A. McLaughlin, and D. M. Hodgson, “Neonatal lipopolysaccharide exposure impairs sexual development and reproductive success in the Wistar rat,” Brain, Behavior, and Immunity, vol. 25, no. 4, pp. 674–684, 2011.
  38. S. Giovanoli, H. Engler, A. Engler et al., “Stress in puberty unmasks latent neuropathological consequences of prenatal immune activation in mice,” Science, vol. 339, no. 6123, pp. 1095–1099, 2013.
  39. K. L. Spalding, O. Bergmann, K. Alkass et al., “Dynamics of hippocampal neurogenesis in adult humans,” Cell, vol. 153, no. 6, pp. 1219–1227, 2013.
  40. A. Tanti and C. Belzung, “Hippocampal neurogenesis: a biomarker for depression or antidepressant effects? Methodological considerations and perspectives for future research,” Cell and Tissue Research, vol. 354, no. 1, pp. 203–219, 2013.
  41. M. Boldrini, M. D. Underwood, R. Hen et al., “Antidepressants increase neural progenitor cells in the human hippocampus,” Neuropsychopharmacology, vol. 34, no. 11, pp. 2376–2389, 2009.
  42. U. Meyer, M. Nyffeler, A. Engler et al., “The time of prenatal immune challenge determines the specificity of inflammation-mediated brain and behavioral pathology,” Journal of Neuroscience, vol. 26, no. 18, pp. 4752–4762, 2006.
  43. M. H. Kaufman, The Anatomical Basis of Mouse Development, Academic Press, San Diego, Calif, USA, 1999.
  44. M. Graciarena, A. M. Depino, and F. J. Pitossi, “Prenatal inflammation impairs adult neurogenesis and memory related behavior through persistent hippocampal TGFβ1 downregulation,” Brain, Behavior, and Immunity, vol. 24, no. 8, pp. 1301–1309, 2010.
  45. T. J. Schwarz, B. Ebert, and D. C. Lie, “Stem cell maintenance in the adult mammalian hippocampus: a matter of signal integration?” Developmental Neurobiology, vol. 72, no. 7, pp. 1006–1015, 2012.
  46. C. M. Teixeira, M. M. Kron, N. Masachs et al., “Cell-autonomous inactivation of the reelin pathway impairs adult neurogenesis in the hippocampus,” Journal of Neuroscience, vol. 32, no. 35, pp. 12051–12065, 2012.
  47. K. Cui, H. Ashdown, G. N. Luheshi, and P. Boksa, “Effects of prenatal immune activation on hippocampal neurogenesis in the rat,” Schizophrenia Research, vol. 113, no. 2-3, pp. 288–297, 2009.
  48. G. Rodríguez-Martínez and I. Velasco, “Activin and TGF-β effects on brain development and neural stem cells,” CNS & Neurological Disorders—Drug Targets, vol. 11, no. 7, pp. 844–855, 2012.
  49. M. Graciarena, V. Roca, P. Mathieu, A. M. Depino, and F. J. Pitossi, “Differential vulnerability of adult neurogenesis by adult and prenatal inflammation: Role of TGF-β1,” Brain, Behavior, and Immunity, vol. 34, pp. 17–28, 2013.
  50. Y. L. Lin and S. Wang, “Prenatal lipopolysaccharide exposure increases depression-like behaviors and reduces hippocampal neurogenesis in adult rats,” Behavioural Brain Research, vol. 259, pp. 24–34, 2014.
  51. K. Järlestedt, A. S. Naylor, J. Dean, H. Hagberg, and C. Mallard, “Decreased survival of newborn neurons in the dorsal hippocampus after neonatal LPS exposure in mice,” Neuroscience, vol. 253, pp. 21–28, 2013.
  52. A. Tanti and C. Belzung, “Neurogenesis along the septo-temporal axis of the hippocampus: are depression and the action of antidepressants region-specific?” Neuroscience, vol. 252, pp. 234–252, 2013.
  53. P. Jiang, Y. Sun, T. Zhu et al., “Endogenous neurogenesis in the hippocampus of developing rat after intrauterine infection,” Brain Research, vol. 1459, pp. 1–14, 2012.
  54. P. Jiang, T. Zhu, W. Zhao et al., “The persistent effects of maternal infection on the offspring’s cognitive performance and rates of hippocampal neurogenesis,” Progress in Neuro-Psychopharmacology & Biological Psychiatry, vol. 44, pp. 279–289, 2013.
  55. S. Farioli-Vecchioli, A. Mattera, L. Micheli et al., “Running rescues defective adult neurogenesis by shortening the length of the cell cycle of neural stem and progenitor cells,” Stem Cells. In press.
  56. G. P. Dias, N. Cavegn, A. Nix et al., “The role of dietary polyphenols on adult hippocampal neurogenesis: molecular mechanisms and behavioural effects on depression and anxiety,” Oxidative Medicine and Cellular Longevity, vol. 2012, Article ID 541971, 18 pages, 2012.

Bruce Hood on the Domesticated Brain (The RSA)

Bruce Hood is the author of The Self Illusion: How the Social Brain Creates Identity (2012). His new book is The Domesticated Brain: A Pelican Introduction, and he was at The RSA in England recently to talk about the new book.

Bruce Hood on the Domesticated Brain

7th May 2014

Listen to the audio  (full recording including audience Q&A)

RSA Replay is now a featured playlist on our Youtube channel, it is the full recording of the event including audience Q&A.
To celebrate the return of Pelican books, Penguin’s groundbreaking and iconic series of intelligent guides to essential topics, we are delighted to announce a new events series bringing together expert minds and curious observers in order to bring vital subjects to life.

Why do we care what others think? What keeps us bound together? How does the brain shape our behaviour?

How did the brain evolve from an organ whose primary function was to help us survive in a threatening world, to an organ which influences our thoughts and behaviour and navigates us through an equally unpredictable social landscape? In the third of these special RSA events, Bruce Hood, award-winning psychologist and director of the Cognitive Development Centre at the University of Bristol will give us a clear and comprehensible insight into the complex mysteries of the brain.

Speaker: Bruce Hood, award-winning psychologist and Director of the Cognitive Development Centre at the University of Bristol.

Chair: Timandra Harkness, writer and performer.

Pelican first appeared in 1937 with the publication of George Bernard Shaw’s ‘The Intelligent Women’s Guide to Socialism, Capitalism, Sovietism and Fascism’ and continued with thousands of books across a massive range of subjects. Aimed at the everyday reader, Pelicans combined intellectual rigour with simple, clear and accessible prose.

Selling over 250 million copies, Pelican in its heyday was seen as influencing the intellectual culture in Britain by lowering the traditional barriers to knowledge. At the time, this confidence in the tastes of the ordinary reader was unusual, and gave Pelican a democratic, populist bent. The first Pelican books cost the same amount as a packet of cigarettes, a radical price at the time, and became especially popular among a self-educating post-war generation.



The Domesticated Brain: A Pelican Introduction by Bruce Hood (Pelican, 2014)

Rudolph Uher - Gene–Environment Interactions in Severe Mental Illness

From Frontiers in Psychiatry: Schizophrenia, this new article looks at the foundations of severe mental illness in gene-environment interactions, which is at least a first step toward grasping the significant impact of environment on all psychological challenges.

Full Citation: 
Uher, R. (2014, May 15). Gene–environment interactions in severe mental illness. Frontiers in Psychiatry: Schizophrenia; 5:48. doi: 10.3389/fpsyt.2014.00048

Gene–environment interactions in severe mental illness

Rudolf Uher [1,2,3]
1. Department of Psychiatry, Dalhousie University, Halifax, NS, Canada
2. Department of Psychology and Neuroscience, Dalhousie University, Halifax, NS, Canada
3. Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, King’s College London, London, UK
Severe mental illness (SMI) is a broad category that includes schizophrenia, bipolar disorder, and severe depression. Both genetic disposition and environmental exposures play important roles in the development of SMI. Multiple lines of evidence suggest that the roles of genetic and environmental factors depend on each other. Gene–environment interactions may underlie the paradox of strong environmental factors for highly heritable disorders, the low estimates of shared environmental influences in twin studies of SMI, and the heritability gap between twin and molecular heritability estimates. Sons and daughters of parents with SMI are more vulnerable to the effects of prenatal and postnatal environmental exposures, suggesting that the expression of genetic liability depends on environment. In the last decade, gene–environment interactions involving specific molecular variants in candidate genes have been identified. Replicated findings include an interaction between a polymorphism in the AKT1 gene and cannabis use in the development of psychosis and an interaction between the length polymorphism of the serotonin transporter gene and childhood maltreatment in the development of persistent depressive disorder. Bipolar disorder has been underinvestigated, with only a single study showing an interaction between a functional polymorphism in the BDNF gene and stressful life events triggering bipolar depressive episodes. The first systematic search for gene–environment interactions has found that a polymorphism in CTNNA3 may sensitize the developing brain to the pathogenic effect of cytomegalovirus in utero, leading to schizophrenia in adulthood. Strategies for genome-wide investigations will likely include coordination between epidemiological and genetic research efforts, systematic assessment of multiple environmental factors in large samples, and prioritization of genetic variants.

Severe Mental Illness

Severe mental illness (SMI) includes the most disabling psychiatric disorders that typically require inpatient treatment, such as schizophrenia, bipolar disorder, and severe depression. Family and molecular genetic studies suggest that schizophrenia, bipolar disorder, and major depressive disorder share common etiology and there may be advantages in studying these disorders jointly (14). This review focuses on these three disorders. Studies of subthreshold psychotic and mood symptoms are also included since they may provide additional information on etiology of SMI.

Both genetic disposition and environmental exposures play important roles in the development of SMI. The risk of SMI runs in families and is shared in proportion to the degree of biological relatedness (5, 6). The overall contribution of genetic factors appears to be stronger for SMI than for common mental disorders (6). Twin studies consistently estimate the heritability of schizophrenia and bipolar disorder in the range of 70–80% (79). The genetic contribution to depression may depend on severity: while general population-based studies find a relatively low heritability around 38% (10), the heritability of hospital-ascertained severe depression was estimated to be between 48 and 75% (11). Molecular genetic studies have recently identified a number of specific genetic polymorphisms that directly contribute to schizophrenia, bipolar disorder, or all types of SMI across populations (1214). The majority of the genetic variants may confer risk to more than one type of mental illness (1, 12).

A number of environmental factors contribute to SMI (Table 1). In utero exposure to infection, lack of nutrients, maternal stress, perinatal complications, social disadvantage, urban upbringing, ethnic minority status, childhood maltreatment, bullying, traumatic events, and cannabis use have all been found to contribute to one or more types of SMI. Some of these exposures appear to be responsible for substantial proportion of cases of SMI. For example, the availability of vitamin D during the prenatal development may be responsible for 44% cases of schizophrenia (15), childhood maltreatment and bullying account for 33% of cases of schizophrenia (16), urban birth and upbringing may be responsible for 35% of cases (17), and use of cannabis in adolescence may account for 14% of cases of schizophrenia (18). A quick addition shows that the above attributable risk percentages sum up to more than 100%. This suggests that multiple factors are likely to contribute to each case of schizophrenia. Some risk factors may be correlated (e.g., a child growing up in urban setting may be more likely to be maltreated) or they may act in synergy (e.g., a person whose early brain development was affected by a lack of vitamin D may be less resilient to the effects of cannabis in adolescence). Nonetheless, the high attributable risks strongly suggest that a significant proportion of cases of SMI may be preventable through modification of environment.


Table 1. Environmental risk factors for severe mental illness.

Gene–Environment Interactions

Gene–environment interactions reflect a causal mechanism where one or more genetic variants and one or more environmental factors contribute to the causation of a condition in the same individual with the genetic factors influencing the sensitivity to environmental exposures (47, 48). They should be distinguished from gene–environment correlations, where genetic factors influence the probability of environmental exposures. Statistically, the likelihood of a gene–environment interaction being present is usually inferred from a significant interaction term between genetic and environmental factor in a multiple regression. Since statistical inference and power depend on the distribution of both the environmental factor and the genetic variant in a particular sample, statistical results often do not correspond to actual biological interaction (49, 50). Therefore, multiple methods of inquiry are required to establish whether a gene–environment interaction is involved (51).

While there are strong environmental risk factors that contribute to a large proportion of cases of SMI, there is also significant evidence of resilience and major individual differences in the impact of environmental exposures (52, 53). Several strong indicators suggest that the marked individual differences in sensitivity to potentially pathogenic exposures are, at least partially, due to genetic factors (54, 55). The combination of very high heritability and strong environmental factors suggests that a large proportion of cases of SMI must be due to a synergy between genetic and environmental causes. If a single environmental factor can explain 30 or 40% of cases of a disease that is 80% heritable, then some of the heritability must be due to joint causation by genes and environment. The way heritability is estimated in twin studies means that gene–environment interactions involving environmental factors that are shared within a family are attributed to the genetic component and contribute to heritability estimates (5557). This misattribution of gene–environment mechanisms to heritability may account for two ostensibly paradoxical observations. First, while some of the strongest known environmental factors (e.g., urbanicity and social disadvantage) are shared within families, twin studies typically estimate no or very small contribution of shared environment (5860). Second, since it has recently become possible to quantify the genetic contribution using molecular genetic data, it became apparent that genetic variants account for much smaller proportion of variance than the twin-based heritability estimates suggested (Figure 1). One of the most likely explanations for the heritability gap is that gene–environment interactions involving shared environmental factors are part of the twin heritability estimates but do not contribute to the molecular heritability estimates that are based on unrelated individuals (4, 55). The large “heritability gaps” for schizophrenia and bipolar disorder suggest that gene–environment interactions may potentially explain a large proportion of cases of SMI.


Figure 1. The heritability gap. Heritability estimates from twin and molecular genetic studies for schizophrenia (SCHZ), bipolar disorder (BPD), and major depressive disorder (MDD) are based on review of twin studies and the results from the Cross-disorder Group of the Psychiatric Genetic Consortium (1, 55). Heritability gap is marked by a blue capped line and quantified as the proportion of total variance in the presence of each disorder. Possible explanations for the heritability gap include gene–environment interactions, inherited rare genetic variants, and overestimation of heritability in twin studies.

Gene–Environment Interactions by Proxy

Several studies have attempted to estimate gene–environment interactions using the familial loading of risk for mental illness as a proxy for genetic factors. A Finnish study has shown that family history of schizophrenia interacts with low birth weight in their effect on educational achievement (61). The link between low birth weight and low educational achievement was much stronger among offspring of biological parents with schizophrenia than in children with no family history of SMI. Since low educational achievement is an antecedent to schizophrenia and major depressive disorder (62), this study may be interpreted as suggesting that gene–environment interactions operate in the early processes on the neurodevelopmental pathway to SMI. This interpretation depends on the assumption that low birth weight is a reflection of environmental factors during pregnancy. However, since a genetic contribution to birth weight is likely (63), the interpretation may become more complex. Several other studies have explored similar proxy gene–environment interactions leading to schizophrenia and other psychotic disorders. A longitudinal Finnish adoption study has shown that excellent parenting and clear communication can substantially reduce the risk of schizophrenia and related conditions among adopted offspring of biological mothers with schizophrenia while no effect of parenting was seen in adopted offspring of biological mothers without SMI (64). Another Finnish study derived data from a population-based registry and showed that serious infection during pregnancy increased the risk of psychosis in offspring who had a family history of psychotic illness (65). A large-scale Swedish adoption study has shown that socio-economic disadvantage during upbringing increased the risk of psychosis in adoptees with a family history of SMI in biological relatives (66). Yet, the pattern is not uniform: a recent study has found a correlation between family history of psychosis and childhood maltreatment (with sons and daughters of parents with psychosis being more often maltreated by their parents), but no interaction between family history of psychosis and childhood maltreatment in the causation of psychotic disorders (67). A twin study of depression found that genetic disposition, indexed by depression in monozygotic and dizygotic co-twins, significantly interacted with environmental triggers (stressful life events) in leading to depressive episodes (68). Taken together, these studies show that pathogenic effects of many but not all environmental risk factors depend on the familial disposition to SMI. Since several of the studies were adoption or twin studies, the familial disposition was separated from the environmental factors and it can be interpreted as a proxy of genetic effects. However, even in adoption studies, there is a residual sharing of environment in the early life and in twin studies monozygotic twins may share more of their environment than dizygotic twins. Consequently, the interpretation of gene–environment studies using proxy measures is limited because familial relatedness cannot be equaled to genetic contribution and because specific environmental factors may interact with specific genetic variants rather than with the multitude of risk alleles that may constitute familial disposition. Therefore, investigation of gene–environment interactions involving specific molecular genetic variants is necessary to advance our knowledge of causal mechanisms leading to SMI.

Gene–Environment Interactions Involving Specific Molecular Genetic Variants

Molecular genetic variants can be measured with high accuracy and their identification may help the development or novel indications for therapeutics. Gene–environment interactions with specific molecular genetic variants have started to be identified in the last decade. Most of the findings have concerned community-ascertained depression or other relatively common mental disorders (69). More recently, several groups of researchers have also investigated and identified specific gene–environment interactions that play a role in the causation of schizophrenia and related conditions (Table 2).


Table 2. Molecular gene–environment interactions in severe mental illness.
The first reported specific gene–environment interaction for a psychotic disorder included a functional polymorphism in the catechol-O-methyltransferase (COMT) gene. COMT codes an enzyme that metabolizes dopamine, the principal neuromediator involved in the positive symptoms of psychosis. A single nucleotide polymorphism (SNP) (rs4680, Val158Met) substitutes valine by methionine (Met) at position 158, leading to the production of an enzyme that is much less efficient than the native Val variant. Caspi and colleagues found that use of cannabis in adolescence led to psychotic symptoms and disorders specifically in individuals carrying the more efficient Val alleles at the functional Val158Met COMT polymorphism (82). While the choice of candidate gene and polymorphism was well justified, the direction of the effect might have been surprising: the more efficient Val allele was associated with sensitivity whilst the less efficient Met allele conferred protection. This finding might have had major implications for personalized prevention of psychosis: a sensitizing genetic variant that explains why many young people remain well even after smoking large amounts of cannabis may help deliver a credible personalized message to those at the highest risk. However, this finding proved difficult to replicate. While initial experimental data supported the interaction (94), several independent studies reported non-replications (83, 9597) or even findings in the opposite direction (98). It appeared that this gene–environment interaction must have been a false-positive finding. However, recent data suggest that there may be a genuine interaction involving COMT and cannabis. Supportive data have been reported from the genetic and psychosis (GAP) study of first onset psychosis together with an explanation why some previous studies might not have found the expected results: the pathogenic effects of cannabis depends on the proportion of tetrahydrocannabinol and cannabidiol (99, 100). When this is taken into account, the gene–environment interaction as reported by Caspi and colleagues was replicated for adolescent exposure to cannabis with high tetrahydrocannabinol to cannabidiol ratio (101). Another refinement has been reported by taking account of childhood maltreatment in addition to the use of cannabis in adolescence: Alemany and colleagues reported a three-way interaction between the COMT genotype Val alleles, childhood maltreatment, and adolescent cannabis use in the etiology of psychotic experiences (84). Most remarkably, this complex three-way interaction was independently replicated by the GROUP investigators: in their sample of Dutch young adults, combination of two COMT Val alleles childhood maltreatment and use of cannabis in adolescence was associated with the highest risk of psychotic experiences (85). While the recent refinements are awaiting further tests, the interim conclusion can be made that COMT and cannabis are likely to be part of a complex causal mechanism leading to psychotic symptoms and schizophrenia.

In the meantime, another genetic polymorphism has been identified that may moderate the effects of cannabis use in development of psychosis. This started with an investigation of 152 genetic variants in 42 selected candidate genes (86). A polymorphism (rs2494732) in the AKT1 gene was identified that interacted with the use of cannabis in the pathogenesis of psychosis: carriers of the C/C genotype on rs2494732 were most likely to develop psychotic illness after smoking cannabis. This interaction is not just highly plausible (AKT1 codes a serine/threonine kinase that relays signal from the cannabinoid receptors), but it appears remarkably robust: the gene–environment interaction between AKT1 rs2494732 and cannabis replicated across three analyses in the primary report (86). Soon after, an independent replication with the same direction of effect was reported in the GAP study of first-episode psychosis patients and healthy controls (87). This effect was driven by daily use of cannabis increasing the risk of psychosis sevenfold in rs2494732 C allele homozygotes, suggesting that avoidance of heavy use of cannabis is highly advisable for individuals carrying this genotype.

A group of Danish researchers focused on another established environmental risk factor for SMI: exposure to virus infection in utero (102104). In one study, they tested 124 SNPs in five genes encoding components of the NMDA glutamatergic receptor, using 365 cases of schizophrenia and 365 matched healthy controls from the Danish population registry (92). They identified two polymorphisms (rs1805539 and rs1806205) in the GRIN2B gene that significantly interacted with maternal positivity for the herpes simplex virus-2 (92). This promising finding is awaiting a replication test.

Additional single candidate gene studies investigated genetic variants with known gene–environment interactions in common mental disorders. The FKBP5 gene coding a co-chaperone of the glucocorticoid receptor was reported to sensitize individuals to developing post-traumatic stress disorder after being exposed to childhood maltreatment (105). Collip and colleagues found a similar gene–environment interaction involving the same SNPs in FKBP5 and childhood maltreatment in increasing the risk of experiencing psychotic symptoms in young adults (90). Perhaps the most investigated gene in relationship to environment is SLC6A4, which encodes the serotonin transporter. A functional length polymorphism in the promoter of SLC6A4, known as 5-hydroxy tryptamine transporter-linked polymorphic region (5-HTTLPR), has been shown to moderate the effects of childhood maltreatment on depression: individuals carrying the short alleles of 5-HTTLPR are prone to develop persistent depressive disorder if they experience maltreatment in childhood (78, 80, 106). Aas and colleagues investigated the interplay between 5-HTTLPR and childhood maltreatment in psychosis and found that the combination of the short 5-HTTLPR alleles and history of childhood maltreatment was associated with cognitive impairment among patients with psychotic disorders (91). This effect was only seen for physical abuse and physical neglect and it did not hold when all types of childhood maltreatment were combined. This finding is waiting for independent replication. A functional polymorphism (Val66Met) in the brain-derived neurotrophic factor (BDNF) gene has been reported to interact with stressful life events and childhood maltreatment in the development of depression, with Met allele carriers being more likely to develop depression after exposure to adversity (71, 107, 108). From a convenience predominantly student sample, Alemany and colleagues have reported that BDNF Met allele carriers with a history of childhood abuse were also more likely to develop psychotic-like experiences (88); this gene–environment interaction has not been replicated in a general population sample of adolescents (89).

Compared to both major depressive disorder and schizophrenia, gene–environment interactions in bipolar disorder have been understudied. Only a single published study has reported that people with bipolar disorder who carried Met alleles at the BDNF Val66Met polymorphism were more likely to develop depressive episodes following stressful life events than Val allele homozygotes (81).

Systematic Search for Gene–Environment Interactions

All the studies reviewed above were restricted to the exploration of one or more polymorphisms in one or more genes that were selected based on their presumed functionality in relation to the disorder or the exposure of interest, i.e., they were candidate gene studies. The study of genetic associations across phenotypes has demonstrated that researchers had not been able to select the right candidate genes: most strong genetic associations turned out to be in genes that no one suspected to be involved (109). In addition, most genetic associations reported from candidate gene studies have proven to be false-positive findings perpetuated through publication bias but not replicated in large-scale systematic studies (110). While candidate gene–environment interactions have had better replicability record (Table 2) (69), the fact that study of gene–environment interactions remains largely limited to functional candidate genes is worrying. It is likely that more cases of SMI can be explained by gene–environment interactions involving genetic variants that no one had suspected than those few polymorphisms explored in the above reviewed studies. Therefore, a systematic search for gene–environment interactions across the genome is the essential next step in establishing the etiology of SMI.

To date, only one systematic search for gene–environment interactions in SMI has been carried out. A group of Danish researchers have searched the genome for genetic variants that may sensitize individuals to developing schizophrenia after being exposed to cytomegalovirus in utero (93). In 488 cases of schizophrenia and 488 healthy controls from the Danish population registry, they measured antibodies to cytomegalovirus in dried blood spots taken from infants at birth (to carry out the Guthrie test for phenylketonuria) and archived. Since the fetus does not produce its own antibodies, these antibodies are of maternal origin and a proxy of maternal infection with cytomegalovirus during pregnancy. From the same dried blood spots, they also extracted DNA and genotyped over half a million SNPs. They did not test interaction with maternal cytomegalovirus infection for all the genotyped SNPs, because of concerns about statistical power. Instead, they carried out a prioritization step and selected 29,000 polymorphisms that were significantly associated with cytomegalovirus infection in the combined case–control sample. This prioritization was based on a proposal that associations between polymorphisms and an exposure in a case–control sample may be induced by a gene–environment interaction (111) (this reasoning is only applicable to case–control samples, and only makes sense for relatively rare disorders). Among the 29,000 SNPs, the rs7902091 in the CTNNA3 gene was found to significantly interact with maternal cytomegalovirus infection in causing schizophrenia after correcting for the number of tests performed. It did not reach the accepted genome-wide level of significance. CTNNA3 encodes a cadherin-associated protein that had been liked to cardiomyopathy, but not suspected to be involved in SMI. It remains to be established whether this gene–environment interaction will prove to be robust in replication.

Future Outlook for Genes and Environment

Adequately powered genome-wide searches for gene–environment interactions should be a priority for the next decade of research. Since the statistical power for detecting gene–environment interactions is lower than statistical power for detecting direct gene-disorder associations (112, 113), large samples will be needed. Paradoxically, these efforts are held back by the unavailability of reliably assessed environmental exposures rather than genome-wide genotyping. The latest genome-wide analyses of schizophrenia, bipolar disorder, and major depressive disorder involved several tens of thousands of cases and tens of thousands of controls each. Yet the largest investigation of gene–environment interactions in schizophrenia involved fewer than 1000 cases. The situation for bipolar disorder is even more striking, with a near absence of gene–environment studies in spite of substantial shared etiology with both depression and schizophrenia and a large heritability gap left to be explained.

The move to systematic genome-wide gene–environment studies will have to overcome major challenges in addition to sample size (69). Several lines of evidence have shown that the quality of assessment of environmental variables is essential. The replicability of interaction between 5-HTTLPR and childhood maltreatment in leading to persistent depression depends on high-quality assessment of childhood maltreatment with detailed interviews or historically recorded variables (114), the replicability of interactions between COMT and cannabis use may depend on how well the exposure to cannabis is characterized, including age and frequency of use as well as the type of cannabis used (101, 115). In the past, large sample collection studies often discounted on the assessment quality, leading to a negative relationship between study size and quality and non-replications in large samples (116, 117). Therefore, when obtaining environmental variables from large samples, substantial efforts will be required to maintain the quality of assessment of environmental variables.

Another challenge lies in the selection of environmental variables to be assessed. A potentially large number of environmental exposures might be contributing to SMI (Table 1). Yet, with each environmental variable added, the number of potential gene–environment tests increases by the number of genetic variables (which is effectively in the range of 500,000–1,000,000 after taking into account linkage disequilibrium between polymorphisms) and the sample size requirements increase accordingly.

Several initiatives have been launched with the aim to collect a systematic selection of environmental variables in addition to genetic material from moderately large samples (118120). Obtaining even larger samples would require a degree of coordination between genetic and epidemiological studies. For example, a funding agency may prioritize funding genotyping only for completely assessed samples with high-quality data on environmental exposures or support genetic sample collections in high-quality epidemiological studies of important environmental exposures. Obtaining genetic and environment data from the same rather than separate samples would create significant opportunities without increasing the total cost of research carried out. With some of these initiatives taking place, our understanding of SMI may substantially evolve over the next decade.

Conflict of Interest Statement
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The Guest Associate Editor Helen Fisher declares that, despite having collaborated with the author Rudolf Uher, the review process was handled objectively.

References are available at the Frontiers site