Fellows Lecture in Pairs: "Prediction and the sensory brain" and "Elucidating regulatory pathways underlying epithelial-to-mesenchymal transition in the developing and adult heart to enable cardiovascular repair and regeneration"

Dr Nicol Harper: “Prediction and the sensory brain”

Sensory neurons are tuned to diverse but distinct features of the sensory input. For example, a particular auditory neuron might respond most strongly to a low-frequency sound that sweeps upwards in frequency, while another might prefer a steady mid-frequency sound. What general principles might explain the diverse tuning properties of neurons in sensory systems? One potential principle is temporal prediction, which proposes that the sensory brain is optimised to represent features in natural stimuli that best predict immediate future inputs. This may help find underlying variables, discard irrelevant information, and better enable future actions. We found that computational model networks optimised for this principle developed tuning properties that resembled those of cortical neurons. This included the oriented spatial tuning of primary visual cortex, the frequency selectivity of primary auditory cortex and, notably, appropriate temporal tuning properties. Furthermore, when such networks were applied to unprocessed audio or video, the resulting tuning properties resembled those found in the cochlea or retina. These networks can be extended hierarchically, a process which produced networks with properties that were consistent with various complex features of neural tuning in primary sensory cortical areas and downstream non-primary cortical areas. The networks can also be made recurrent, which resulted in network connectivity that resembled synaptic connectivity patterns of sensory cortical areas. Finally, these models use prediction and prediction error to learn, and we have evidence from electrophysiology of an unusual form of neural representation of such signals in cortex. These results together suggest that many aspects of sensory processing may be optimised to extract those features with the most capacity to predict future input.

Nicol Harper Biography:

Nicol Harper completed his PhD in optimal coding in the auditory system at University College London with David McAlpine, with some time in the lab of Shihab Shamma at the University of Maryland, College Park. After this he secured a Sir Henry Wellcome Postdoctoral Fellowship, splitting his time between the Redwood Centre for Theoretical Neuroscience lead by Bruno Olshausen at the University of California, Berkeley, and the Auditory Neuroscience Group lead by Andrew King at the University of Oxford. He is now a University Research Lecturer in the Auditory Neuroscience Group where he researches normative principles and stimulus encoding in sensory neural systems.

Dr Joaquim Vieira: “Elucidating regulatory pathways underlying epithelial-to-mesenchymal transition in the developing and adult heart to enable cardiovascular repair and regeneration”

My group uses mouse genetics, human embryonic stem cells, and cutting-edge molecular and imaging approaches to gain mechanistic insight into normal heart development. We are particularly interested in understanding cell and molecular interactions at the epicardium-myocardium interface that support growth and maturation of cardiac muscle, as well as development of the coronary vasculature, formation of the cardiac lymphatic network, and patterning of the cardiac autonomic nervous system in the subepicardial space. Epithelial-to-mesenchymal transition (EMT) of cells from the epicardium (outermost layer of the heart) is essential for these processes to take place in the embryo and, therefore, holds great therapeutic potential in the context of heart disease where new muscle and vessels are warranted. Yet, the molecular underpinnings of epicardial EMT and subepicardial cell-cell crosstalk remain poorly understood. Here, we discuss our current research exploring the functional requirements for cis-regulatory sequences in the Wilms’ tumour 1 (Wt1) locus which encodes for the key EMT regulator WT1 and elusive antisense long noncoding transcripts (Wt1as lncRNA) during heart development and disease (myocardial infarction). In addition, we discuss ongoing unbiased screens integrating CUT&RUN, scRNA- and ATAC-sequencing to identify WT1-downstream targets, as well as further regulators of epicardial EMT. Taken together, a better understanding of the cellular and molecular mechanisms supporting mammalian heart development is poised to identify molecular targets to effect (adult) heart repair and regeneration.

Research funded by the British Heart Foundation (Intermediate Basic Science Research Fellowship, Oxford BHF Centre of Research Excellence) and John Fell Fund.

Joaquim Vieira Biography:

Joaquim Vieira studied Biochemistry as an undergraduate at the University of Lisbon, Portugal, and researched developmental angiogenesis for his PhD in Professor Christiana Ruhrberg’s laboratory at University College London (UCL). After successful postdoctoral studies at UCL and University of Oxford with Professor Paul Riley investigating the epigenetic regulation of the epicardium and role of cardiac lymphatics during heart development and disease, he was awarded a British Heart Foundation (BHF) Intermediate Basic Science Research Fellowship to start his research group in Oxford. His group studies the role of long noncoding RNAs and enhancers during epithelial-to-mesenchymal transition in the developing and diseased heart; and retains an interest in understanding cell-cell interactions in the developing heart supporting coronary and lymphatic vasculature expansion, as well as patterning of the cardiac autonomic system. A better understanding of the cellular and molecular mechanisms supporting organ development is poised to identify molecular targets to effect (adult) heart repair.