La família, el centre de l’activitat de cuidar: parentification role

The functional architecture of the C. elegans motor circuit is highly analogous to that of more complex invertebrate and vertebrate models

In Chapter 2, I showed that the C. elegans circuit for forward locomotion contains multiple subunits capable of independent rhythm generation. These units, whether isolated by surgical or optogenetic manipulations, invariably showed higher frequency oscillations in the posterior of the animal than in the anterior. I further showed that extensive, bidirectional rhythmic coupling exists within the circuit and that the animal’s normal frequency of locomotion in our assay was generally intermediate between the frequencies of head and tail oscillations isolated by optogenetic relaxation of the neck muscles. These findings suggest that rhythmic forward locomotion is not driven by one oscillator overruling and entraining the others, but by the network as a whole operating as a large rhythmic unit with properties that may be different from any individual unit in isolation.

Despite the fact that C. elegans is a non-segmented, millimeter-long animal with only 300 neurons, this functional architecture for locomotion is highly analogous to that shown by more complex species. Leech swimming is driven by central pattern generators located in each of the 18 ganglia along the body segments (Kristan et al., 2005; Mullins et al., 2011). As shown in Figure 4.1, a gradient exists among the rhythm geniting units in isolation, with the lowest frequency oscillators found closest to the head and the highest frequency closer to the mid-body. Moreover, the oscillatory frequency of the intact chain of oscillators is always intermediate between the frequencies of the

121

highest and lowest frequency ganglia in isolation (Zheng et al., 2007). Coordination between the oscillatory units to allow them to synchronize and reach an intermediate frequency requires extensive, bidirectional inter-oscillator coupling, and indeed such a mechanism exists in the leech (Weeks, 1981), just as I found that it exists in C. elegans.

The analogy between motor function in C. elegans and other animals even extends to vertebrates. The swimming lamprey and zebrafish also contain distributed rhythm generating circuits along the segments of the spinal cord, with their own gradients in excitability and rhythmic properties, but linked by substantial intersegmental coupling (Kiehn, 2006; Mullins et al., 2011).

Taken together, my findings suggest that the functional architecture of the C. elegans

forward motor circuit is highly analogous to those found in other vertebrate and invertebrate models.

122

Figure 4.1. Computational modeling of the leech swim CPG.

Reproduced with modifications from (Zheng et al., 2007). Upper panel: periods of the CPGs within the body ganglia in isolation show a gradient from head to tail. Lower panel: the overall swim CPG in the intact circuit is intermediate between the highest and lowest frequency ganglia.

123 Mechanisms of rhythm generation

The extensive functional similarity between C. elegans and more complex animals offers the hope of understanding a highly conserved neuronal strategy in deeper detail than has been possible in more complex model organisms. Because of the unique

experimental tractability of C. elegans in genetics and neuroscience, it is possible to investigate rhythm generating circuits at the network, cellular, and molecular levels to understand the neuronal basis of locomotion.

The first objective of future studies should be to discover the network-level mechanisms that allow the whole-body rhythmic unit to generate locomotory rhythms. Historically, there has been a great debate over whether rhythms arise from central pattern

generators intrinsic to the nervous system, which can generate oscillations even without the physical execution of movement, or from sensory reflex loops that drive oscillations by detecting and reacting to each body bend (Kristan et al., 2005; Mullins et al., 2011) (Figure 4.2 A, B). Although central pattern generators have been demonstrated in many organisms, it is also well appreciated that even where they exist, sensory feedback is highly capable of entraining central clocks, suggesting that its contribution to rhythmic movement cannot be ignored (Figure 4.2C). For example, injection of sinusoidal current into leech stretch sensitive VSR neurons was found to completely entrain rhythmic activity in the VNC motor circuits (Yu et al., 1999; Yu and Friesen, 2004).

124

Figure 4.2. Overview of proposed methods of rhythm generation.

(A) A central pattern generator (CPG, green circle) is capable of generating rhythmic signals even if the execution of movement, the bending of the muscles (tan ovals), is interrupted. (B) A reflex loop generates oscillations concomitant with the execution of body bends. Stretch sensors on each side sense bending of the muscle and then activate motor neurons to stimulate bending in the opposite direction. This creates a reflex loop that drives oscillations. (C) The leech contains a central pattern generator, but injection of sinusoidal current into the VSR stretch sensitive neurons can entrain the CPG and define the locomotory behavior.

125

To determine which mechanism(s) explain C. elegans forward locomotion, Chris Fang- Yen and I have designed new experiments that were performed by Pilar Alvarez-Illera and Shelly Teng. We briefly interrupted normally progressing forward locomotion by hyperpolarizing neurons or muscles for a fraction of a second. If the rhythm generator is not attuned to sensory feedback – a blind central pattern generator – the rhythm should continue in phase after only a brief disruption (Figure 4.3A). If, on the other hand, sensory reflexes play a key role in rhythm generation or modulation, a brief interruption of forward movement should permanently shift the phase of locomotion after the manipulation. Our results (not shown), strongly support the latter possibility.

Moreover, I have attempted to search for C. elegans CPGs by conducting calcium recordings of the B-type motor neurons in mechanically immobilized worms. These experiments were performed with the help of Alice Liu. Our results showed no clear pattern of oscillation in these motor neurons, except for some rhythmic activities in two of the anterior-most B motor neurons. These results do not rule out the possibility of CPGs existing along the remainder of the motor circuit, but as of yet no evidence for that possibility has emerged.

Taken together, my preliminary results show that proprioception is closely associated with the generation of body rhythms, but do not support or disprove the idea that CPGs such as cellular pacemakers exist within the circuitry. Hence, much more work needs to be done before we will have a clear picture of how the motor circuit generates rhythms. The starting point for these studies will likely be additional calcium imaging of the motor neurons combined with simultaneous optogenetic stimulation of motor neurons or interneurons.

126

Figure 4.3. Identifying a CPG or a reflex loop by its response to brief inhibition of movement.

(A) An internal CPG that runs without concern for the execution of body movements would be only transiently effected by the manipulation, and then resume its ongoing course. (B, C) A reflex loop – or a CPG that is highly sensitive to the execution of movement – would be permanently delayed by the inhibition, leading to a phase shift.

127

Depending on where in the cycle the inhibition occurred, the phase shift may be an advance or a delay.

128

Towards an integrated network, cellular, and molecular understanding of rhythmic locomotion

As described in Chapters 1 and 2, substantial progress towards understanding the network and cellular principles that govern rhythmic locomotion has been made in other organisms such as the leech and the lamprey. Although this highly fruitful work can and should continue, it must be appreciated that not only are the pictures of rhythm

generation in these animals incomplete, but integrating these findings with their associated molecular mechanisms is restricted by the lack of genetic access to these species. Even genetically accessible model organisms such as the mouse and zebrafish also do not provide an ideal platform for attaining an integrated understanding of

locomotion because their spinal networks are fantastically complex, and genetic manipulations in longer life-cycle animals are extremely costly.

C. elegans, with its 3 day life cycle, genetic manipulability, and amenability to rapid genetic screening, provides the most tractable platform to develop a multilevel understanding of rhythmic locomotion. The most obvious targets for genetic efforts should be screens to identify the molecules involved in stretch sensation in the B motor neurons, which have yet to be found (Wen et al., 2012). Not only would characterization of these molecules help explain how wave propagation and communication between oscillatory units is achieved, but the preliminary results described above suggest that stretch sensitive molecules may play a key role in generating rhythmic bends in C.

elegans. A genetic screen for these molecules can be approached by looking for animals

with defective tail movement during forward swimming. Because of the high degree of homology between the C. elegans genome and that of humans (Lai et al., 2000), it is also quite possible that the molecular machinery uncovered may have a role in human

129

movement as well. The ability to dissect these molecular components and pathways, and study them in the context of individual cells or small networks of neurons, makes C.

elegans perhaps the most promising model organism available today for a multilevel

130

APPENDIX I: QUANTITATIVE ASSESSMENT OF FAT LEVELS IN