Pierce JT et al. (1998) West Coast Worm Meeting "A NEW MECHANISM FOR CHEMOTAXIS IN C. ELEGANS"

Lockery SR, Pierce JT

[West Coast Worm Meeting, 1998]

Little is known about the behavioral mechanism of chemotaxis in C. elegans. The close proximity of its chemosensory organs implies that the worm samples chemical concentration at a single point in space. Thus, C. elegans chemotaxis is likely to be a response to the rate of change of attractant concentration (dC/dt). We tested this idea using a tracking system to record the instantaneous position, speed, and turning rate of single worms (N = 77) as a function of time during chemotaxis in Gaussian-shaped gradients of ammonium chloride or biotin. Analysis of the distribution of turning rates showed that each worm track could be divided into periods of smooth swimming (runs) and periods of frequent turning (pirouettes). The occurrence of pirouettes was strongly correlated with dC/dt, but not with absolute concentration (C) or its second derivative (d2C/dt2). Pirouettes were most likely when the worm was heading down the gradient (dC/dt < 0), and were virtually absent when the worm was heading up the gradient (dC/dt > 0). Further analysis revealed that on average the direction of movement following a pirouette was up the gradient. These observations suggest that chemotaxis is produced by a series of course-correcting pirouettes triggered by movement down the gradient. We tested the course-correction model in two ways. First, we imposed the correlation between pirouettes and dC/dt < 0 on a simulation of worm motion derived from empirical distributions of speed and turning rate during normal locomotion. The simulation reproduced the behavior of real worms in Gaussian-shaped gradients and also novel, planar gradients. Thus, the course-correction model of C. elegans chemotaxis is sufficient and general. Second, we asked whether the course-correction model could account for the behavior of mutants with impaired chemotaxis. Tracking unc-23(e611) during chemotaxis, we found that while pirouettes were triggered normally, the average direction of movement after a pirouette was the same as before it. This circumscribed deficit is consistent with the unc-23 phenotype in which head movements are limited by muscle defects but the chemosensory system is spared. Thus the course-correction model accounts for the chemotaxis behavior of normal and mutant worms. We are currently using neuronal ablations and mutants with nervous-system defects to identify the neural basis of the course correction model of chemotaxis in C. elegans.

TM Morse et al. (1999) International C. elegans Meeting "Time-series analysis of pirouettes in C. elegans chemotaxis."

TM Morse, JT Pierce, SR Lockery, ML Moravec, TC Ferree, SE Owens, AB Bates

[International C. elegans Meeting, 1999]

Chemotaxis in C. elegans involves a series of abrupt turns (pirouettes) triggered by movement down a gradient of chemical attractant * . Analysis of the time series of concentration change experienced by a worm, together with its pirouette record, suggests a three-stage chemotaxis mechanism in which chemical concentration (C(t)) is differentiated (dC(t)/dt), smoothed (q(t)), and converted into pirouette probability by a nonlinear function (P(q[t])). To test the plausibility of this mechanism, we constructed a computer model in which the smoothing filter and the nonlinearity were estimated from the time series of dC(t)/dt and the pirouette records of real worms. Pirouettes were modeled by sampling randomly, via a Poisson process with probability P(q[t]), from the distribution of direction changes associated with pirouettes in real worms. The average chemotaxis index (time-average of normalized C(t)) of model worms (0.47 +/- 0.05 SD, n = 2000) closely matched the average chemotaxis index of real worms (0.45 +/- 0.04 SD, n = 45), indicating that the three-stage mechanism is quantitatively sufficient to account for C. elegans chemotaxis. To determine the form of the smoothing filter and nonlinearity directly, we have devised a new behavioral assay that subjects unrestrained worms to negative-going impulses in dC/dt as they swim across a sharp border between high and low concentrations of attractant. Preliminary results show that immediately after a border crossing, large impulses make P(t)~= 1, while small impulses make 0 t) >< C. elegans . * Pierce, J.T., and Lockery, S.R., J. Neurosci. (Submitted)

Nathan A Dunn et al. (2003) International Worm Meeting "A Survey of Neural Network Models for Chemotaxis in C. elegans"

Shawn R Lockery, Nathan A Dunn, John S Conery

[International Worm Meeting, 2003]

C. elegans chemotaxis involves bouts of turning (pirouettes) modulated by the rate of change of attractant concentration. A candidate neural network for chemotaxis in C. elegans is emerging several ablation studies. The aim of this research is to generate testable models of how the chemotaxis network computes the sensorimotor transformation. Neurons in the models were passive, isopotential electrical compartments that mimicked likely biophysical properties of C. elegans neurons. Networks began with ten neurons: a sensory neuron representing the chemosensory neurons, an output neuron representing the state of the locomotory command neurons, and eight generic interneurons. Neurons were fully interconnected with self-connections. The input to the sensory neuron was attractant concentration C(t) recorded from a real worm during a chemotaxis assay. Pirouette probability P(t) in the model was determined according to the activity state of the output neuron. Networks were optimized by adjusting neuronal time constants and connection strengths via an interactive search algorithm (simulated annealing) until P(t) was high when dC/dt<0 and P(t) was low when dC/dt>0, as in real worms. After optimization we eliminated interneurons that did not have a significant effect on P(t); 63% of our networks could be reduced to one interneuron. One-interneuron networks had three common features. (1) A differentiation circuit involving a rapid, excitatory, monosynaptic pathway from the input neuron to the output neuron in parallel with a delayed, inhibitory disynaptic pathway via the interneuron. Thus, the activity of the output neuron reflected the difference between present and previous values of C(t), an approximation to dC/dt. (2) Strong inhibitory self-connections on the input and output neurons, but a weak inhibitory self-connection on the interneuron. Mathematical analysis of self-connections showed that they decreased the effective time constant of the neurons. We tested this by training a set of networks with a delay between C(t) and desired pirouette probability. The magnitude of the inhibitory self-connections varied inversely to the delay, as expected. (3) Recurrent inhibitory connections between all pairs of neurons. Analysis of the recurrent connections suggested that they stabilize activity of neurons. Anatomical substrates for all three features have been found in previous reconstructions of the C. elegans nervous system; our results provide hypothetical functions for these patterns of connectivity. We are currently analyzing the common features of the multi-interneuron networks. Support: NSF IBN0080068.

Tafelmeyer, Petra et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "High complexity domain libraries for yeast two-hybrid analysis of protein interactions in C. elegans"

Formstecher, Etienne, Tafelmeyer, Petra, Masson, Maryline, Pintard, Lionel, Boulin, Thomas, Bessereau, Jean-Louis

[C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany, 2010]

Yeast two-hybrid (Y2H) protein interaction screening has proven instrumental for the analysis of the C. elegans interactome, thanks to dedicated resources such as ORF collections and oligo dT-primed cDNA libraries. However, map completeness has been limited so far by the use of full-length proteins (ORFeome) or C-terminal polypeptide fragments (oligo dT-primed cDNA libraries) that resulted in significant false negative rates. To circumvent these limitations, we have used a domain-based strategy to construct two highly complex, random-primed C. elegans cDNA libraries. The first library has been constructed by combining equimolar amounts of mRNA from 4 different N2 samples to increase transcript diversity: (1) males and hermaphrodites (all stages including embryos), (2) starved mixed stage culture, (3) heat-shocked mixed-stage culture, (4) dauer stage. The second library was prepared exclusively from C. elegans embryos of all stages. The complexity of both libraries is greater than 10 million independent fragments in yeast, with an average fragment size of 800 bp. To ensure reproducible and exhaustive Y2H results, these libraries are screened at saturation using an optimized mating procedure. This allows to test on average 100 million interactions per screen, corresponding to a 10-fold coverage of the library. As a consequence, multiple, independent fragments are isolated for each interactant, enabling the immediate delineation of a minimal interacting domain and the computation of a confidence score. These two C. elegans libraries have been integrated into our high-throughput yeast two-hybrid platform and are available for screening on a fee-for-service basis. The results of representative screens performed on both libraries will be presented at the meeting.

Yamazoe, Akiko et al. (2013) International Worm Meeting "A neuronal mechanism for navigation along a repulsive odor gradient."

Iwasaki, Yuishi, Hashimoto, Koichi, Kawazoe, Yuya, Fujita, Kosuke, Busch, Karl Emanuel, Iino, Yuichi, Gengyo-Ando, Keiko, Nakai, Junichi, Fei, Xianfeng, Yamazaki, Shuhei, Tanimoto, Yuki, Miyanishi, Yosuke, Yamazoe, Akiko, Kimura, Kotaro

[International Worm Meeting, 2013]

For survival and reproduction, animals navigate toward or away from certain stimuli, which requires the coordinated transformation of sensory information into motor responses. In worms, the pirouette and the weathervane strategies are considered the primary navigation strategies for responding chemosensory stimuli. We found, however, that worms use a novel navigation strategy in odor avoidance behavior: In a gradient of the repulsive odor 2-nonanone, worms efficiently avoid the odor, and ~80% of initiation of long, straight migrations ("runs") were away from the odor source, which cannot be simply explained by the two known major strategies. Direct measurement of local odor concentration suggested that pirouettes are efficiently switched to runs when worms sense negative dC/dt of 2-nonanone. To test whether runs are indeed caused by negative dC/dt, we established an integrated microscope system that tracks a freely moving worm during stimulation with a virtual odor gradient and simultaneously allows for calcium imaging and optogenetic manipulations of neuronal activity (Tanimoto et al., this meeting). Using this system, we found that a realistic temporal decrement in 2-nonanone concentration (~ 10 nM/sec) caused straight migration by suppressing turns. We also found that a pair of AWB sensory neurons were continuously activated during the odor decrement and that optogenetic activation or inactivation of AWB neurons suppressed or increased turning frequency, respectively. In addition, we found that ASH nociceptive neurons increased turning frequency during odor increment. Taken together, our data indicate that the counteracting turn-inducing and turn-suppressing sensory pathways can effectively transform temporal sensory information into spatial movement to select the right path leading away from potential hazards.

Thomas Boulin et al. (2007) International Worm Meeting "High complexity, random-primed, domain libraries for yeast two-hybrid analysis of C. elegans interactome."

Thomas Boulin, Sebastien Dubleumortier, Laurent Daviet, Etienne Formstecher, Jean-Louis BESSEREAU

[International Worm Meeting, 2007]

Yeast two-hybrid (Y2H) protein interaction screening has proven instrumental for the analysis of C. elegans interactome, thanks to dedicated resources such as ORF collections and oligo dT-primed cDNA libraries1,2. However, map completeness has been limited so far by the use of full-length proteins (ORFeome) or C-terminal polypeptide fragments (oligo dT-primed cDNA libraries) that resulted in significant false negative rates3. To circumvent these limitations, we have used a domain-based strategy to construct two highly complex, random-primed cDNA libraries. The first library has been constructed by combining equimolar amounts of mRNA from 4 different N2 samples to increase transcript diversity: (1) males and hermaphrodites (all stages including embryos), (2) starved mixed stage culture, (3) heat-shocked mixed-stage culture, (4) dauer stage. The second library was prepared exclusively from C. elegans embryos of all stages. The complexity of both libraries is greater than 10 million independent fragments, with an average fragment size of 800 bp. To ensure reproducible and exhaustive Y2H results, these libraries are screened at saturation using an optimized mating procedure. This allows to test on average 100 million interactions per screen, corresponding to a 10-fold coverage of the library. As a consequence, multiple, independent fragments are isolated for each interactant, enabling the immediate delineation of a minimal interacting domain and the computation of a confidence score4. These two C. elegans libraries have been integrated into our high-throughput yeast two-hybrid platform and are available for screening on a fee-for-service basis. The results of representative screens performed on both libraries will be presented at the meeting. Please send inquiries to: Etienne Formstecher, eformstecher@hybrigenics.com 1. Zhang et al., 2004, Nature Genetics, 36:507 2. Li et al., 2004, Science, 303:540 3. Han et al., 2005, Nature Biotechnology, 23:839 4. Formstecher et al., 2005, Genome Research, 15:376.

Blaxter ML et al. (1995) International C. elegans Meeting "Expressed sequence tags from the parasitic nematode Brugia malayi identify otherwise anonymous C. elegans genes"

Blaxter ML, Raghavan N, Gosh I, Maizels RM, Scott AL

[International C. elegans Meeting, 1995]

The C .elegans genome sequencing initiatives have identified many genes which have no similarity to other database sequences. Given the deep evolutionary separation between nematodes and other model organisms we suspect that many of these will be nematode-specific. We have initiated an est project on the spirurid parasite B.=A0malayi (a causative agent of human lymphatic filariasis): we aim to identify genes important in the development of infective L3 larvae in the invertebrate/vertebrate host transition. A full length library was made in the lambda ZapII vector from SL-oligo(dT) cDNA from infective L3. So far 300 clones have been 5' tag sequenced. These can be grouped into 4 classes: (1) Clones unique to our dataset; (2) universal housekeeping genes; (3) homologues of functionally identified nematode genes (both C. elegans and parasites) and (4) clear homolgues of otherwise anonymous C. elegans ORFs and DNAs. Classes (3) and (4) constitute a possible "nematode-specific" group and the availability of such phylogenetically distant homologues should aid both parasitic and freeliving research communities.

Yamazoe, Akiko et al. (2011) International Worm Meeting "Worms maintain their response to the repulsive odor 2-nonanone during avoidance behavior."

Yamazoe, Akiko, Kimura, Kotaro, Iwasaki, Yuishi

[International Worm Meeting, 2011]

Animals can maintain their behavioral response to environmental stimuli even under unstable environmental conditions and during various animal movements. To investigate neural mechanisms for such robust behavioral responses, it is necessary to quantitatively analyze the time-course changes in the correlation between the stimulus and behavioral response. For this, we quantitatively analyzed stimulus as well as behavior of worms' avoidance response to repulsive odor 2-nonanone. When animals migrate away from a source of repulsive signal, their avoidance response is likely weakened. In a previous study, however, we have shown that worms exhibited a constant average velocity of avoidance from 2-nonanone for 10 min (Kimura et al., J. Neurosci., 2010), suggesting a neural mechanism for such constant avoidance. In addition to the quantitative analysis of avoidance response to 2-nonanone (Yamazoe & Kimura, CeNeuro, 2010), we recently developed a technique to measure the concentration of 2-nonanone at specific spatial and temporal points of gas phase in the assay plate. By using a highly sensitive gas chromatograph, we observed a clear gradient of 2-nonanone, of which concentration increased with time. Based on this measured gradient of 2-nonanone, we determined the 2-nonanone concentration that each worm experienced during the avoidance assay (Cworm) and observed the following: (1) During the first 2 min of the assay worms did not initiate avoidance response and migrated randomly, and Cworm increased continuously up to the order of mM at 2 min. (2) After 2 min, worms started to migrate farther away from the odor source, and Cworm was maintained around the concentration, despite increase in the concentration gradient. (3) Cworm decreased effectively during runs, while it increased and decreased largely during pirouettes. (4) When compared between the early and late phases of the assay, the maximum dCworm/dt in each run decreased several fold along with the avoidance behavior, even though the orientation directions did not change considerably; that is, even when the gradient of 2-nonanone became shallower, the accuracy of worm orientation appeared maintained. These results suggest that worms may increase sensitivity to dC/dt during exposure to a certain concentration of 2-nonanone. We are currently conducting computer simulation to test this hypothesis. Further analysis may help us uncover the mechanism of maintaining proper behavioral responses.

V Kapulkin et al. (2001) International C. elegans Meeting "The worm is not enough - comparative genomics of Ancylostoma ceylanicum"

A Ciesielski, A Magalska, H Wedrychowicz, V Kapulkin

[International C. elegans Meeting, 2001]

Ancylostoma ceylanicum is a parasitic nematode which infects humans and domesticated animals. We have taken a comparative genomic approach to understanding A. ceylanicum parasitism by constructing cDNA and genomic DNA libraries from this parasitic nematode. We build cDNA libraries using directional, SL1- oligo dT, RT-PCR amplification, followed by topoisomerase I- based A/T cloning into a pCR-XL-TOPO (Invitrogen). PCR products were size selected by horizontal gel slicing and fractions were subsequenty separately cloned to build sub- libraries. In our view this strategy allows rapidly identification of highly complex sub- libraries and therefore eliminate fractions containing highly redundant inserts. Using similar strategy we build a small insert genomic library. Briefly, genomic DNA was sheared, gel fractionated and cloned into pCR4Blunt-TOPO (Invitrogen). Both cDNA and gDNA libraries were therefore build in high-copy plasmid based vectors useful for direct sequencing. To characterize our libraries we sequenced over one hundred clones. We find that about half of SL-cDNA clones show apparent homology to genes predicted by C. elegans genome project. Results of trial sequencing and characteristics of sub- libraries will be summarized.

Williams, Kyle C. et al. (2011) International Worm Meeting "Essential Functions of IFA-2 Domains in C. elegans Fibrous Organelles."

Plenefisch, John, Williams, Kyle C.

[International Worm Meeting, 2011]

The C. elegans fibrous organelle (FO) complex consists of apical and basal hemidesmosomes linked by cytoplasmic intermediate filaments (IFs). FOs are found most prominently in the epidermis where they overlie body wall muscle and serve to transmit force from the muscle to the cuticle. IFA-2, one of four epidermally expressed IFs whose loss results in epidermal fragility and failure of muscle-cuticle force transmission, localizes to FOs. Unlike IFB-1 and IFA-3 that are required embryonically, IFA-2, although expressed in the embryo, is not essential for FO function until postembryonic stages. This distinctive phenotype allows us to specifically explore the functions of IFA-2 domains, and map their interactions with other FO associated proteins. All IFs contain three domains: a central rod domain and globular head and tail domains. The rod is essential for assembly of IFs into mature filaments while less is known about the functions of the head and tail domains. Roles of the head and tail domains of IFA-2 were examined by expressing GFP-tagged IFA-2 variants in transgenic animals and examining their phenotype and localization in wild-type and null backgrounds. Mutant variants included IFA-2 deleted for the head domain (DH), deleted for the tail domain (DT), deleted for both (RO), and variants that contained only the head domain (HO) or tail domain (TO). The expression and function of DH is virtually indistinguishable from full-length IFA-2. DT results in a lowered incorporation of IFA-2 into FOs and is unable to rescue a null allele of ifa-2, suggesting the tail domain is essential to IFA-2 function. Though both HO and TO are expressed well in the hypodermis at all stages, incorporation into FOs is variable between individual animals. High levels of early-stage HO incorporation into FOs correlate with a dominant-negative muscle detachment phenotype, while low-level incorporation results in healthy animals. This suggests the head domain interacts with FO components essential to tissue integrity. To determine the protein components of the FOs that interact with the head and tail domains, yeast two-hybrid and co-localization experiments are in progress. This work should help to elucidate the role of IFA-2 in FO function and reveal the underlying defects leading to tissue fragility.