PhD Thesis
Abstract
Repeated social stress experienced during pregnancy âprogrammesâ long-term changes in the offspringâs brain and behaviour which may prepare the offspring for an adverse postnatal environment. However, a âmismatchâ between the predicted and actual postnatal environment can result in maladaptation to stress. The effects of prenatal stress (PNS) are sexually dimorphic â for example, male, but not female PNS offspring display heightened anxiety-like behaviour, whereas hyperactivity of the hypothalamo-pituitary-adrenal (HPA) axis in response to stress is reported in both sexes. The mechanisms underlying the (mal)adaptive programming PNS induces are unclear, however, may result from changes to the offspringâs central GABAergic system, thereby impacting the inhibitory regulation GABA exerts on the HPA axis and central limbic areas that regulate anxiety-like behaviour. Moreover, recent advances in understanding the gut-brain axis provide an additional avenue of investigation regarding the long-term effects of PNS on other systems. The following key questions were addressed in this thesis: (1) Does PNS lead to resilient or vulnerable stress-related phenotypes when faced with additional chronic stress in adulthood; (2) is central GABAA receptor (GABAAR) subunit expression altered by PNS?; and (3) does PNS with or without additional chronic stress alter the gut microbiome or predispose offspring to dysbiosis? To address these questions, PNS offspring were generated by exposing pregnant dams to social stress on gestational days 16-20.
Here, PNS alone did not have any significant adverse effect on anxiety-like behaviour in the offspring, regardless of sex, however chronic stress (seven days variable stressors) in adulthood lead to reduced anxiety-like behaviour in both male and female PNS offspring. Moreover, there was a significant difference in plasma corticosterone concentrations following anxiety behavioural testing, with PNS females displaying reduced corticosterone secretion, regardless of chronic stress exposure; while male offspring exposed to chronic stress exhibited higher corticosterone concentrations. In addition, testosterone concentrations were significantly elevated by chronic stress alone. In females, estradiol was also significantly reduced by chronic stress, regardless of PNS status. The results suggest that the animals in this study may be exhibiting resiliency to adverse conditions (i.e. chronic stress) experienced in later life, which appears to be independent of HPA-axis activity.
In males, the GABAAR âș1 receptor subunit expression in the hippocampus was affected by both PNS and chronic stress (down-regulated). Whereas in the prefrontal cortex (PFC), there was an interaction between PNS and CS, wherein this subunit did not exhibit the PNS-only increase in expression. A similar effect was also seen for âș5 in the PVN of males, while an interaction prevented the PNS-only decrease in âș5 expression in the PFC. In females, PNS had significant effects on increasing the expression of âș1 in the hippocampus and amygdala (CeA, MeA), as well as the expression of âș5 in the PVN, hippocampus, and amygdala. Chronic stress had significant effects on âș1 expression in the hippocampus (decreased expression), âș2 expression in PFC (decreased), and âș5 in the PVN (increased). Finally, there was an interaction between PNS and CS increasing the expression of âș1 in the amygdala (BLA, BMA). Overall, the results suggest that PNS and chronic stress may influence inhibitory GABAergic control in key brain regions involved in mediating anxiety-like behaviour and regulating HPA axis activity.
In pregnant dams, the gut microbiome was not affected by social stress and the offspringâs microbiota were significantly different to those of the pregnant dams, suggesting that any dysbiosis in the adult offspring is independent of maternal gut microbiome changes. Analysis of fecal samples taken from offspring exposed to chronic stress, as well as antibiotic-gavage-treated offspring, reveal that these treatments had a greater effect on microbiome composition than PNS. However, PNS males showed a moderate sensitivity to antibiotic treatment compared to controls, with significant changes to bacterial families of the orders Clostridiales, Bacteroidales and Enterobacteriales. Indeed, while PNS had a moderate effect on the microbial composition of the gut in males, chronic stress-induced alterations in the abundance of specific microbial communities in both males and females. Nevertheless, neither antibiotic nor water gavage treatment had a significant effect on anxiety-like behaviour, suggesting that, at least for this model, the gut microbiome does not play a role in mediating PNS phenotypes.
In conclusion, the results suggest that adverse PNS effects may have potentially been tempered by underlying stress resiliency with respect to anxious behaviour. The central expression of GABAAR subunits is differentially sensitive to PNS and chronic stress, and these changes may contribute to altered behaviour and stress sensitivity. Finally, PNS alone has little effect on the composition of the gut microbiome, which is evidently more sensitive to chronic stress in adulthood, than stress exposure during prenatal life. However, PNS males treated with antibiotics showed a sensitivity to gut perturbations. Overall, these findings contribute to our understanding of the maladaptive effects of PNS and the mechanisms underlying PNS outcomes.
Lay Summary
Stress experienced during pregnancy can have life-long negative repercussions for the offspring. This is considered an evolutionary trait to prepare the offspring for an adverse postnatal environment, as experienced by the pregnant mother. However, in the absence of a challenging postnatal environment, this prenatal âprogrammingâ can have harmful effects on the offspring, leading to mood disorders, such as anxiety or depression, and abnormal responses to stress. In healthy individuals, stress activates the stress-response axis (known as the hypothalamo-pituitary-adrenal axis; HPA axis) allowing the body to respond appropriately to the challenging stimulus, for example by ensuring energy is available, and then return to a non-stressed baseline. While the HPA axis is maintained by several checkpoints, dysfunction of this regulation can occur. This leads to an excess of stress hormones (cortisol in humans, corticosterone in rats) which can have damaging effects and is commonly associated with anxiety and depressed mood. Prenatal stress (PNS) can lead to abnormal responses to stress: for example, male, but not female PNS offspring display heightened anxious behaviour, whereas hyperactivity of the HPA axis in response to stress is reported in both sexes. It is currently not known how exactly this programming occurs, or indeed the full scope of the changes in the brain that lead to anxious behaviour, however, they may result from changes to one of the checkpoints involved in maintaining control of the HPA axis.
Rapid communication between neurons occurs through electrical impulses. In order to generate these impulses, positively and negatively charged ions move in and out of nerve cells through channels, which are gated by receptors. When a neurotransmitter attaches to its receptor, the channel opens and ions flow through. These neurotransmitters can therefore be excitatory (opening the channel to positively charged ions that make the neuron more likely to produce an impulse) or inhibitory (opening the channel to negatively charged ions that make the neuron less likely to fire). It has been suggested that an imbalance between excitatory and inhibitory neurotransmission may underlie the abnormal stress responses seen in PNS rats. The primary inhibitory neurotransmitter in the brain is called GABA. It binds to GABAA receptors, preventing the neuron from firing and thereby passing on an electrical impulse. The GABAA receptor is one of the checkpoints regulating the HPA axis and preventing HPA axis hyperactivity. GABAA receptors are composed of multiple building blocks (subunits), which can each affect receptor activity and function. For example, the subunits âș1, âș2, and âș5 have been associated with anxiety and genetic deletion studies have demonstrated that any of these subunits can either promote or prevent anxiety, depending on the brain area from which they are removed.
In order to investigate how prenatal stress programmes offspring, rats were used to model these changes by exposing pregnant rats to social stress at the end of pregnancy, and then raising the pups until adulthood, at which point testing was carried out. Here, PNS alone did not lead to an increase in anxiety behaviour, regardless of sex, however chronic stress in adulthood did lead to reduced anxiety behaviour. Moreover, there was a significant difference in corticosterone concentration in the blood, with PNS females displaying reduced corticosterone, regardless of chronic stress exposure; while male offspring exposed to chronic stress exhibited higher corticosterone concentrations than male offspring not exposed to chronic stress. These results suggest that the rodents in this study may be exhibiting resiliency to adverse post-natal conditions (i.e. chronic stress) which appears to be independent of HPA-axis activity.
Further, while chronic stress alone had a small impact on GABAA receptor subunits, PNS or a combination of PNS and chronic stress changed where and how GABAA receptor subunits were expressed in a sex- and brain-region dependent manner. Given that the stress response is mounted via these regions, PNS-dependent alterations, especially in the face of chronic stress in adulthood, may offer insight into the mechanisms that led to an increase in stress resiliency.
In addition to investigating changes in GABAA receptor subunits in the brain, this thesis also examined the role of the gut microbiome in mediating the programming effects of PNS. The gut microbiome describes the communities of bacteria that naturally colonise the gut and play an important role in regulating brain chemistry. Recent advances in understanding this gut-brain interaction have demonstrated an association between the gut microbiome and stress responses. The results suggest that PNS only moderately altered the bacterial composition of the gut microbiome and the offspring microbiome was not âinheritedâ from their stressed mothers. However, in males and females, PNS and chronic stress exposure increased pro-resiliency bacteria. In addition, PNS may lead to altered sensitivity to antibiotic treatment, as there were changes in bacterial communities in control offspring that were not apparent in PNS offspring treated with antibiotics.
Overall, the results of this thesis suggest that 1. males and females exhibited heightened stress-resiliency when faced with an adult (chronic stress) challenge, 2. PNS-alone or in combination with chronic stress led to adaptations of GABAA receptors in stress-sensitive brain regions, and 3. In PNS offspring, the gut microbiome may have been sensitive to disruption by chronic stress (possibly improving stress-resiliency), and antibiotic treatment (in males).
Taken together, these results provide new avenues of investigation to continue to understand foundational mechanisms underlying stress responsivity and anxiety, which in the future may allow for a more target approach to the treatment in humans.
Reflection
After reading this thesis, this final section may seem unconventional and perhaps unnecessary; At the beginning of it, my acknowledgements thank all the wonderful people (and creatures) that have supported me throughout this long, long period in my life. It seems appropriate, if a little indulgent, that a glimpse into the rear-view mirror of my PhD somewhat bookends my thesis. Something that I think often goes unacknowledged is everything that the PhD teaches you. For many students starting out in their PhDs, it might seem impossible and unattainable to produce a final thesis. I certainly had my moments. Others, at the end of their journey, might forget to look back and take stock of everything that they have learned, as if that knowledge had always been there. I think that does the PhD a disservice; We are here to learn, grow, and develop as researchers. Often through trial and error, and more error. I would like to take this short section to acknowledge all the things I didnât know or wasnât as skilled in, before I started, but I have now learned how to do. After all, a PhD is an incredible opportunity to learn and develop the skills needed to succeed in the future.
Practical Skills
Before starting this PhD, I had no appreciable experience in in vivo animal maintenance, handling, or the general research skills required to work with rodents, including administering treatment by gavage or behavioural testing. The months spent in the animal unit helped me learn all those skills. The same goes for all the in vitro experiments carried out in this PhD: never before had I done tissue sectioning, a radioimmunoassay, radionucleotide-labelled in situ hybridisation assays, or 16S rRNA sequencing. These are all skills that I can now say that I have learned and learned from. I have learned from the trials and the errors. I have learned from the successes and from attempts that never even made it to a finished experiment. Finally, I have learned from writing up the thesis and from my viva. Without mistakes, nothing can be learned.
Computational and Statistical Skills
What I learned in this section came as a big, beautiful surprise to me. Learning how to use R had been a goal of mine when starting the PhD; I knew that undergraduates were given the opportunity to learn this and it felt like an incredibly powerful tool that I wanted to become at least familiar with. I never imagined how far that goal would take me. Starting with a simple 10-line R script made available to me by Dr Crispin Jordan, I took each piece apart to learn what it did. I then added more lines, and took even more away. I slowly learned how the language worked and became familiar with how to use it effectively. For someone with a computational background, learning R may sound terribly dull, as it is in fact a very intuitive language, but for me, this was the first time I had undertaken to learn any type of coding language. The most surprising thing about this however, was that I fell in love. I loved the logic puzzles new problems would present. I loved that I could just try something and see what happened. I could get lost in solving these and took pride in succeeding to find a solution. I learned how to carry out complex analysis, picking up new skills in statistical analysis and theory, as well as how to create the figures that I wanted.
It was a good thing that I became so delighted by coding, as I soon discovered that the initial pre-processing steps required for the microbiome analysis required me to code in the command-line interface using a different language, Bash. If you are computer scientist, I will remind you again at this point that I am not. For anyone wondering: this is the equivalent of saying I learned how to eat with a knife and fork, and feeling quite pleased about it. But in this analogy, I didnât even know cutlery existed before trying to use it. So I set to learning what I needed to know and managed to navigate my way through the bioinformatics portion of the microbiome analysis. Even then, when I was able to return this prepared data to my beloved R, I had to learn how to apply specific microbiome-analysis commands and processes. So I did. I tried and I met hurdles and impasses. I made mistakes. I sought help on forums and from colleagues, until I succeeded.
When I moved on to the image analysis step of my thesis, which was not without its own set of problems, I realised that many of these steps could be automated and thereby made more efficient and reliable. So I set about learning the complex functionalities of ImageJ and QuPath, eventually learning how to write macros that knitted these steps together in a streamlined process in order to create a new analysis pipeline that can be used by others. The section of Chapter 4 in this thesis about the development of the pipeline is an ode to the joy and pride I felt developing it. From scratch. By myself. I think if I told pre-PhD Danny that I would end up creating something like this, she wouldnât believe it.
Finally, and this may not be obvious to the reader of this formatted pdf document, I wrote my thesis in LaTeX. I had no prior knowledge of this particular language when I sat down to start writing, but it seemed like a good idea to learn. So I did. More trial and more error ensued, and I think you will agree that the outcome looks good. It was certainly preferable to negotiating with a certain common text processor about image placement.
Research Skills
Everything I mentioned so far has been a very practical, hands-on skill. However, I think it is also worth mentioning the more inexplicit, but important research skills I learned. The first one being experimental design, both from reading papers and learning from the experiments I carried out, but in all honesty, a portion of this skill has also come from the 20:20 hindsight of my own work, wishing I had done things slightly differently. Now I know better. I think this a common lament amongst PhD students. Additional skills that I have learned and will carry with me, and hopefully continue to improve on are data management, scientific writing, presentation skills, and critical scientific evaluation. I know so much more about these than when I started, but I donât think learning ever ends for skills like these. Finally, and I think I have made this point clearly enough throughout so I wonât belabour it any more: independent problem solving.
Professional Skills
The PhD also taught me incredibly valuable skills in collaboration, asking colleagues for advice, adapting to challenges and meeting those challenges with a resiliency I did not know I possessed. That is not to say I have not had my set-backs or my moments of despair. But I learned to get back up and try again tomorrow. And finally, I learned what my personal strengths and weaknesses are. I learned what I am passionate about and how that passion can fuel my creativity to become a better scientist, and remain as ever, a life- long learner. In closing, it has been a privilege to learn how to be a scientist.