Increased Transmitter Release and Aberrant Synapse Morphology in a Drosophila Calmodulin Mutant

Corresponding author: Michael Stern, Department of Biochemistry and Cell Biology, 6100 S. Main Street, Rice University, Houston, TX 77005-1892., stern{at}bioc.rice.edu (E-mail).

Communicating editor: R. S. HAWLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The ubiquitous calcium-binding protein calmodulin (CaM) has been implicated in the development and function of the nervous system in a variety of eukaryotic organisms. We have generated mutations in the single Drosophila Calmodulin (Cam) gene and examined the effects of these mutations on behavior, synaptic transmission at the larval neuromuscular junction, and structure of the larval motor nerve terminal. Flies hemizygous for Cam^3c1, a mutation in the first Ca²⁺-binding site, exhibit behavioral, neurophysiological, and neuroanatomical abnormalities. In particular, adults exhibit defects in locomotion, coordination, and flight. Larvae exhibit increased neurotransmitter release from the motor nerve terminal at low [Ca²⁺] in the presence of the K⁺ channel-blocking drug quinidine. In addition, synaptic bouton structure at motor nerve terminals is altered. These effects are distinct from those produced by altering the activity of the CaM target enzymes CaM-activated kinase II (CaMKII) and CaM-activated adenylyl cyclase (CaMAC). Furthermore, previous in vitro studies of mutant Cam^3c1 demonstrated that although its Ca²⁺ affinity is decreased, Cam^3c1 protein can activate CaMKII, CaMAC, and CaM-activated phosphatase calcineurin in a manner similar to wild-type CaM. Thus, the Cam^3c1 mutation might affect Ca²⁺ buffering or interfere with the activation or inhibition of a CaM target distinct from CaMKII or CaMAC.

CALMODULIN (CaM) is a ubiquitous, highly conserved, Ca²⁺-binding protein that mediates many cellular signaling processes (reviewed by COHEN and KLEE 1988 ), including those of the nervous system, where very high levels of CaM and several of its targets are found (KLEE and VANAMAN 1982 ; KLEE 1991 ; HANSON and SCHULMAN 1992 ). Two prominent neuronal CaM targets, Ca²⁺/CaM-dependent protein kinase II (CaMKII) and CaM-activated adenylyl cyclase (CaMAC), have been investigated extensively and implicated in murine behavior and learning (SILVA et al. 1992 ; CHEN et al. 1994 ; BUTLER et al. 1995 ), cultured neurite outgrowth (SOLEM et al. 1995 ; WILLIAMS et al. 1995 ), and associative learning in Aplysia (ABRAMS et al. 1991 ) and Drosophila (DUDAI et al. 1984 ; GRIFFITH et al. 1993 ). Other neuronal CaM targets include the CaM-activated phosphatase calcineurin (NICHOLS et al. 1994 ), the ryanodine receptor Ca²⁺ channel of brain and muscle (GUERRINI et al. 1995 ), and the GAP-43 protein (STRITTMATTER et al. 1995 ). Finally, in the Paramecium, CaM controls cellular excitability via effects on Na⁺ and K⁺ channels (reviewed by PRESTON et al. 1991 ).

In Drosophila, CaM is encoded by a single gene (Cam; SMITH et al. 1987 ) and is amenable to genetic analysis; however, this approach is complicated because Camnull mutants die at the first larval instar stage, with behavioral defects including a high frequency of backwards movement (HEIMAN et al. 1996 ). The reduction of the activity of CaM or a CaM target by induced expression of a peptide inhibitor provides one useful approach for the genetic dissection of CaM function (GRIFFITH et al. 1993 ; VANBERKUM and GOODMAN 1995 ). An alternative approach is to generate point mutations affecting specific amino acids within CaM in the hope that these specific alterations might perturb the interaction of CaM with distinct target enzymes. This analysis has been performed on CaM in yeast and Paramecium, and has revealed that single or double amino acid substitutions can each affect a distinct subset of cellular processes (OHYA and BOTSTEIN 1994A , OHYA and BOTSTEIN 1994B ; reviewed by PRESTON et al. 1991 ).

To begin a genetic dissection of the roles of Drosophila CaM in nervous system function and development, we recently generated several point mutations in Cam (NELSON et al. 1997 ). One point mutation of particular interest, termed Cam^3c1, is a glutamate-to-lysine substitution at amino acid 31 in the first Ca²⁺ binding site, as illustrated in Figure 1. This mutant protein had previously been generated in vitro, and it was found that despite significantly altered Ca²⁺-binding properties (MAUNE et al. 1992A , MAUNE et al. 1992B ), this mutant CaM functioned similarly to the wild-type in the activation of several CaM targets (GAO et al. 1993 ; GUPTAROY et al. 1996 ).

View larger version (27K): In this window In a new window Download PPT slide   Figure 1. A diagram representing the Drosophila CaM protein, with the change caused by the Cam3c1 mutation indicated. The roman numerals I, II, III, and IV represent the four Ca2+-binding sites.

Here we show that Cam^3c1/Cam^null individuals display abnormal adult behavior, as well as abnormal transmitter release and morphology at the larval neuromuscular junction. The defects we observe for Cam^3c1/Cam^null are different from those exhibited by flies with reduced CaMKII or CaMAC function. Thus, the Cam^3c1 mutation might affect Ca²⁺ buffering or interactions between CaM and a target distinct from CaMKII or CaMAC.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Mutants and stocks:
The isolation of Cam^null has been described previously (HEIMAN et al. 1996 ). The isolation of Cam^3c1 is described in NELSON et al. 1997 . The isogenic parent chromosome of Cam^null, called Cam², contains a P element 34 bp upstream of the transcription start site of Cam (HEIMAN et al. 1996 ). Flies homozygous for this insertion are viable and fertile. The parent chromosome of Cam^3c1 was marked with the eye color marker brown (bw) and isogenized before mutagenesis. CyO P[y⁺] is a derivative of the CyO balancer chromosome that carries the marker y⁺. All other genes, mutants, or special chromosomes not described in this section can be found in LINDSLEY and ZIMM 1992 .

Sequence analysis of Cam^3c1:
RNA was prepared from Cam^3c1/Cam^null flies with Trizol (GIBCO BRL, Gaithersburg, MD) according to the manufacturer's instructions. Double-stranded cDNA was prepared using SuperScript II reverse transcriptase (GIBCO BRL) according to the manufacturer's instructions, except that random hexanucleotides were used to prime the first strand at 42°. PCR fragments were generated using primers to the 5' and 3' untranslated regions of Cam. PCR products were subcloned into pBluescript (Stratagene, La Jolla, CA) for DNA sequencing. Two independently generated PCR fragments were sequenced, and both exhibited the same mutation.

Behavioral tests:
Flies of the desired genotypes were crossed at 21–22° in uncrowded bottles. Flies from each bottle were etherized and collected twice, 2 days apart. The first collection was on the second day after the initial eclosions in the bottles. After collection, flies were rested for 2 days in individual vials before testing. Four behavioral tests were performed essentially as described previously (RICHARDS et al. 1996 ) and as described in the Table 1 legend.

Electrophysiology:
Larval dissections, muscle recordings, and muscle voltage clamping were performed as described previously (JAN and JAN 1976 ; GANETZKY and WU 1982 ; STERN and GANETZKY 1992 ). Saline solution used for dissections and recordings was 5 mM HEPES, pH 7.1, 128 mM NaCl, 2 mM KCl, 4 mM MgCl₂, 35.5 mM sucrose, and CaCl₂, as specified in the text. Because Cam^null is a recessive lethal allele, it was not possible to generate lines homozygous for this mutation. Thus, to obtain larvae carrying Cam^null, we crossed the y mutation into all stocks that were used for electrophysiology and anatomy experiments, and we used the balancer CyO P[y⁺] to enable the identification of larvae carrying the Cam alleles of interest by their yellow mouthhooks.

Muscle cell 6 from abdominal segments 4 or 5 were used for data collection. Quinidine (Sigma, St. Louis, MO), if used, was bath applied from a 20 mM stock solution immediately before experimentation. The wild-type larvae described were obtained from a cross from the parental stocks of each Cam allele (described above). Data were collected, digitized, and analyzed with the Superscope and MacAdios or instruNet systems from GW Instruments.

Neuroanatomy:
Wandering third instar larvae of the appropriate genotypes were identified by their yellow mouthhooks as described above. Larvae were dissected, and immunohistochemistry was performed essentially as described previously (BUDNIK et al. 1990 ). Fixation was performed in paraformaldehyde fixative (4% paraformaldehyde in 0.1 M phosphate buffer) for 3–4 hr. Preparations were washed three times for 15 min each in 0.1 M phosphate buffer containing 0.1% Tween-20 (PBT) and incubated overnight in 1:200 anti-horseradish peroxidase antibody (anti-HRP; Jackson ImmunoResearch Labs., West Grove, PA) diluted in PBT. After three 15-min washes in PBT, preparations were incubated for 4 hr in 1:20 goat anti-rabbit HRP-conjugated IgG (Cappel/Organon Teknika, Durham, NC), diluted in PBT, and washed twice in PBT and once in 0.05 M Tris buffer for 15 min each. The preparations were then reacted using diaminobenzidine (Sigma) and 0.3% hydrogen peroxide (EM Science). The stained synaptic terminal structures were visualized with Nomarski optics on a Nikon Optiphot-2.

Transformation rescue:
The Cam cDNA was cloned downstream of the promoter hsp70 and introduced into flies by P-element-mediated germ-line transformation, as described elsewhere (NELSON et al. 1997 ). One transformant carried hsp70-Cam on the X chromosome. Males carrying hsp70-Cam and Cam^3c1 balanced with CyO[y⁺] were crossed to females lacking hsp70-Cam and carrying Cam^null balanced with CyO[y⁺]. After daily 1 hr heat shocks at 37°, Cam^3c1/Cam^null larvae or adults were selected and assayed phenotypically as described above. Female progeny from this cross contained one copy of hsp70-Cam and thus served as the rescued progeny, whereas male progeny from this cross lacked hsp70-Cam and thus served as the nonrescued internal controls.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Sequence analysis (see MATERIALS AND METHODS) established that Cam^3c1 is a G-to-A transition in codon 31 that changes the encoded amino acid from glutamate to lysine (Figure 1). Cam^null is a deletion mutation that removes exons 0 and I of Cam and confers a transcript-null phenotype (HEIMAN et al. 1996 ). As controls for all of our experiments, we have used larvae or adult flies transheterozygous for the two Cam⁺ parent chromosomes of these Cam mutations: bw-3 (the parent chromosome of Cam^3c1, marked with the eye color mutation bw) and Cam² (the parent chromosome of Cam^null). Cam²/Cam^null and bw-3/Cam^null larvae, or larvae homozygous for either Cam² or bw-3, each exhibit normal synaptic transmission at the larval neuromuscular junction and normal morphology of the motor nerve terminal (data not shown). Thus, in the electrophysiological and anatomical assays that we used, both Cam² and bw-3 larvae appear similar to other wild-type larvae that have been reported. To test for possible dominant phenotypes associated with either the Cam^3c1 or Cam^null chromosomes, experiments were also performed on flies or larvae carrying either Cam^3c1 or Cam^null heterozygous with the wild-type parental chromosome of the other allele (Cam² or bw-3, respectively).

Behavioral defects of Cam^3c1/Cam^null:
The Cam^3c1 mutation was isolated on the basis of viability defects in combination with Cam^null (NELSON et al. 1997 ). We found that Cam^3c1/Cam^null flies exhibit reduced locomotion, coordination, and flight ability when compared to Cam heterozygous or wild-type control flies (Table 1). These behavioral defects, which are qualitatively similar to those reported earlier by NELSON et al. 1997 , are consistent with a defect in neuromuscular function or development in this Cam mutant combination. Cam^3c1/+ and Cam^null/+ flies performed somewhat more poorly than wild-type control flies in these behavioral assays (Table 1). This result suggests that each Cam mutation tested could produce a certain dominant behavioral phenotype, or that the accumulation of additional dominant mutations induced on the second chromosome during mutageneses and unlinked to Cam could generate these behavioral phenotypes. In each test, however, Cam^3c1/Cam^null mutant flies performed substantially more poorly than flies of either heterozygous genotype, which suggests that the behavioral defects exhibited by Cam flies result mostly or entirely from defects at Cam.

Transmitter release at the Cam^3c1/Cam^null larval neuromuscular junction:
We used the larval neuromuscular preparation (JAN and JAN 1976 ) to determine if Cam^3c1/Cam^null third instar larvae exhibited any defects in transmitter release at the neuromuscular junction. Muscle currents [termed excitatory junctional currents (ejcs)] evoked by nerve stimulation and transmitter release were monitored with the muscle held under voltage clamp. We found no significant differences in ejc amplitude between Cam^3c1/Cam^null and wild-type control larvae at any of the four [Ca²⁺] tested (Figure 2).

View larger version (17K): In this window In a new window Download PPT slide   Figure 2. Neurotransmitter release at the neuromuscular junction in Cam3c1/Camnull larvae. (A and B) Voltage clamp recordings of wild-type control (Cam2/bw-3) and Cam3c1/Camnull mutant ejcs in response to nerve stimulation at [Ca2+] of 0.15 and 0.4 mM, respectively. The traces shown are averages of 10–15 traces from a single larva. (C) Mean amplitudes of ejcs from wild-type and Cam3c1/Camnull mutant larvae at the indicated [Ca2+]. Average amplitudes for 10–15 traces from each of six larvae were pooled to determine the means and SEMs (indicated by error bars). Holding potential was -60 mV.

The effects on motor neuron function and transmitter release of many behavioral mutations, particularly those affecting ion channels, are enhanced by application of the K⁺ channel-blocking drug quinidine. This drug, when applied at a concentration of 0.1 mM, completely and specifically blocks the delayed rectifier K⁺current in the Drosophila larval muscle (SINGH and WU 1989 ). Application of quinidine enhances the effects of ion channel mutations, such as Shaker and Hyperkinetic, on the duration of motor nerve terminal depolarization and transmitter release (WU et al. 1989 ). The phenotypes of other excitability mutants, such as inebriated and pushover, are also enhanced by quinidine application (STERN and GANETZKY 1992 ; RICHARDS et al. 1996 ). We were interested in determining if quinidine might affect transmitter release in Cam^3c1/Cam^null as well. As shown in Figure 3, application of quinidine had no significant effect on ejc amplitude in the wild-type control larvae or in larvae heterozygous for either Cam mutation. In contrast, we found that quinidine application to Cam^3c1/Cam^null at the three lowest external [Ca²⁺] tested caused an approximately threefold increase in ejc amplitude. These effects were statistically significant (P < 0.01 for [Ca²⁺] = 0.1 mM, P < 0.001 for [Ca²⁺] = 0.15 and 0.2 mM). Quinidine application to Cam^3c1/Cam^null at 0.4 mM [Ca²⁺] caused no significant change in ejc amplitude.

View larger version (16K): In this window In a new window Download PPT slide   Figure 3. Neurotransmitter release at the neuromuscular junction in Cam3c1/Camnull larvae in the presence of quinidine. (A and B) Voltage clamp recordings of ejcs from wild-type control (Cam2/bw-3), Cam heterozygous, and Cam3c1/Camnull larvae in the presence of 0.1 mM quinidine in response to nerve stimulation at an external [Ca2+] of 0.15 and 0.4 mM, respectively. The traces shown are an average of 10–15 traces taken from a single larva of each genotype. (C) Mean amplitudes of ejcs from wild-type control, heterozygous control, and Cam3c1/Camnull mutant larvae at the indicated [Ca2+]. Average amplitudes for 10–15 traces from each of six larvae were pooled to determine means and SEMs (indicated by error bars). *P < 0.01. **P < 0.001 by Student's t-test. Holding potential was -60 mV.

Effects of Cam^3c1/Cam^null on spontaneous transmitter release:
The increased amplitude ejc described in Figure 3 for Cam^3c1/Cam^null larvae probably resulted from increased release of transmitter from the motor nerve terminal or from increased sensitivity of the muscle to transmitter. These possibilities were distinguished by examining the amplitude of miniature excitatory junctional currents (mejcs) in the muscle, which result from the spontaneous release of single vesicles containing transmitter. We tested mejc amplitudes under six different conditions: in the presence or absence of quinidine at each of three different [Ca²⁺] (0.1, 0.15, and 0.2 mM). We found no significant differences in mejc amplitude among wild-type, Cam/+, or Cam^3c1/Cam^null larvae under any condition tested. Table 2 shows results for each of the four genotypes in the presence of 0.1 mM quinidine and 0.15 mM [Ca²⁺], which is the condition for which the Cam^3c1/Cam^null exhibits its strongest effects on evoked transmitter release. These observations suggest that the muscle responsiveness to neurotransmitter is normal, and that the increased amplitude ejcs observed in Cam^3c1/Cam^null larvae reflect increased transmitter release.

Certain mutations that affect evoked transmitter release also affect the frequency of spontaneous transmitter release. These mutations include kinesin heavy chain, synaptotagmin, and cysteine string protein (GHO et al. 1992 ; LITTLETON et al. 1993 ; UMBACH et al. 1994 ). We found that Cam^3c1/Cam^null larvae exhibited a normal frequency of mejcs, suggesting normal spontaneous transmitter release under each condition tested. Table 2 shows mejc frequency at an external [Ca²⁺] of 0.15 mM and in the presence of 0.1 mM quinidine.

Defects in synapse morphology of Cam^3c1/Cam^null:
Mutations in several genes that affect transmitter release at the neuromuscular junction also affect axon branching or the number of synaptic boutons (BUDNIK et al. 1990 ; ZHONG et al. 1992 ). In addition, inhibition of the CaM target CaMKII, which causes increased transmitter release, also causes increases in axon branching (WANG et al. 1994 ). To determine if Cam^3c1/Cam^null larvae also exhibited defects in the number or structure of synaptic boutons, we used antibodies against HRP to visualize the motor nerve terminal. This visualization revealed structural abnormalities in the nerve terminals of Cam^3c1/Cam^null larvae in muscle 13 of abdominal segments 3–5 (Figure 4 and Table 3). In particular, rather than cascading into a string of distinct type I and II boutons as in the control larvae, the terminal arbor of the Cam^3c1/Cam^null larvae forms a thickened, or large, misshapen structure with few distinct boutons. The large structure results in a reduced number of boutons and a nearly complete lack of terminal branching in muscle 13. In comparison to control larvae, no abnormalities in the structure of nerve terminals on muscles 6, 7, or 12 have been observed. Despite this altered synapse morphology, muscle 13 synaptic transmission in Cam^3c1/Cam^null larvae resembles the muscle 6 properties that we have described. This observation raises the possibility that this defective bouton might still be functional.

View larger version (88K): In this window In a new window Download PPT slide   Figure 4. Synaptic terminal structure at the neuromuscular junction. Larval longitudinal muscle 13 of abdominal segment A4 was used in both a wild-type control larva and in the Cam3c1/Camnull mutant. The nerve terminals were visualized with antibodies against HRP. (A) Wild-type hemisegment showing the wild-type pattern of innervation, as indicated by arrow. (B) Cam3c1/Camnull mutant hemisegment showing the "typical" abnormal innervation pattern (arrow). (C) Cam3c1/Camnull mutant hemisegment showing the "extreme" abnormal innervation pattern (arrow).

Phenotypes of Cam^3c1/Cam^3c1:
We were interested in determining if the defects in the neuromuscular junction that we observed for Cam^3c1/Cam^null larvae also occurred in larvae homozygous for Cam^3c1. We found that Cam^3c1/Cam^3c1 larvae exhibited normal synaptic transmission, even in the presence of quinidine, and normal structure of synaptic boutons at the neuromuscular junction (data not shown). These results indicate that the phenotypic defects we have described require a 50% reduction, rather than normal levels, of the Cam^3c1 mutant protein.

Phenotypic rescue by ectopic expression of Cam:
To determine if expression of wild-type Cam under control of the heat shock promoter could rescue the phenotypes conferred by Cam^3c1/Cam^null, we created Cam^3c1/Cam^null flies harboring the hs-Cam⁺ transgene and assayed these flies as described in MATERIALS AND METHODS. Control Cam^3c1/Cam^null flies (denoted hs-control) lacked the hs-Cam⁺ transgene, but were heat shocked in parallel with the rescued flies. Rescued and control larvae and adults were compared for behavioral, electrophysiological, and anatomical properties.

Behavioral tests: As shown in Table 1, Cam^3c1/Cam^null hs-Cam⁺ flies exhibited behaviors that were comparable to flies heterozygous for Cam^3c1 or Cam^null. In contrast, hs-control flies exhibited behaviors similar to Cam^3c1/Cam^null, except that hs-control flies exhibited an apparent partial rescue of the climbing defect. This apparent rescue might be associated with the reduced viability of Cam^3c1/Cam^null flies after heat shock that we observed (data not shown): increased lethality of Cam^3c1/Cam^null as a result of the heat shock might selectively target the more uncoordinated variants in the population, leaving the stronger variants available for the behavioral assay.

Electrophysiological tests: Transmitter release in rescued vs. hs-control larvae was assayed at an extracellular [Ca²⁺] of 0.15 mM in the presence of 0.1 mM [quinidine]. For these rescue assays, transmitter release was monitored as an extracellular junctional potential (ejp), which is the muscle depolarization elicited by nerve stimulation and transmitter release from the motor neuron nerve terminal. As shown in Figure 5, hs-control Cam^3c1/Cam^null larvae exhibited a 2.7-fold increase in ejp amplitude compared to Cam^3c1/Cam^null larvae expressing the hs-Cam⁺ transgene (6.9 ± 1.0 vs. 2.6 ± 1.0 mV, respectively, P < 0.05), indicating that expression of a Cam⁺ transgene could restore transmitter release to normal levels. The ejps in these hs-control and rescued larvae were very similar in amplitude to ejps recorded from Cam^3c1/Cam^null and wild-type control larvae (8.8 ± 3.0 and 2.7 ± 1.1 mV, respectively, P < 0.01), as shown in Figure 5.

View larger version (19K): In this window In a new window Download PPT slide   Figure 5. Rescue of the neurotransmitter release defect in Cam3c1/Camnull larvae by ectopic expression of Cam+. (A) The ejps in wild-type control (Cam2/bw-3), Cam3c1/Camnull, Cam3c1/Camnull hs-Cam+, and Cam3c1/Camnull hs-control in the presence of 0.1 mM quinidine in response to nerve stimulation at an external [Ca2+] of 0.15 mM. Cam3c1/Camnull hs-Cam+ and Cam3c1/Camnull hs-control larvae were heat shocked every day before electrophysiological recordings. (B) Mean amplitudes of ejps from wild-type control, Cam3c1/Camnull, Cam3c1/Camnull hs-Cam+, and Cam3c1/Camnull hs-control larvae. For wild-type control, n = 9; for Cam3c1/Camnull, n = 13; for Cam3c1/Camnull hs-Cam+, n = 8; and for Cam3c1/Camnull hs-control, n = 8. Means and SEMs (indicated by error bars) are shown. *P < 0.05, **P < 0.01.

Anatomical tests: Synaptic bouton structure was measured in rescued Cam^3c1/Cam^null and in Cam^3c1/Cam^null hs-control larvae (Table 3). We found that hs-control larvae exhibited a high frequency of the thickened, misshapen bouton we have reported, demonstrating that daily heat shocks alone did not detectably affect this phenotype. In contrast, this phenotype was rescued in larvae expressing the hs-Cam⁺ transgene: thickened, misshapen boutons were not observed in these larvae.

The observation that each of the three phenotypes observed in Cam^3c1/Cam^null flies could be rescued by ectopic expression of Cam⁺ provides additional evidence that these phenotypes result specifically from defects at Cam.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have investigated the effects of a mutation in Cam, called Cam^3c1, which results from a substitution of glutamate to lysine in codon 31, on adult behavior and on function and development of the larval motor neuron. We found that adults hemizygous for Cam^3c1 exhibit defects in locomotor behavior, coordination, and flight. Furthermore, larvae hemizygous for Cam^3c1 exhibit aberrant transmitter release and synaptic bouton structure at the larval neuromuscular junction. Phenotypes similar to these have not been previously observed and do not resemble those of previously characterized changes in CaM target genes.

What defects in CaM activity might result from the Cam^3c1 mutation? The Cam^3c1 mutant protein had previously been generated in vitro and the effects of the mutation upon Ca²⁺ binding, Ca²⁺-induced conformational change, and target activation have been examined (MARTIN et al. 1992 ; MAUNE et al. 1992A , MAUNE et al. 1992B ; GAO et al. 1993 ; MUKHERJEA and BECKINGHAM 1993 ; GUPTAROY et al. 1996 ). This mutation, termed B1K, affects a glutamate in the first Ca²⁺-binding site, which is similarly positioned in each Ca²⁺-binding site and is essential for binding Ca²⁺ at each site. This mutant CaM is capable of normally activating several target enzymes, including mammalian CaMAC and calcineurin (MAUNE 1991 ; GAO et al. 1993 ) and seven isoforms of the Drosophila CaMKII (GUPTAROY et al. 1996 ). These findings, along with computer modeling studies, have led to the hypothesis that the charged amino group of the lysine substitution might elicit the normal Ca²⁺-induced conformational change associated with Ca²⁺ binding at site 1 (MAUNE et al. 1992A , MAUNE et al. 1992B ). Recent NMR studies (A. CRIVICI and M. IKURA, personal communication), however, suggest that the conformation of Ca²⁺-binding site 1 in B1K is not wild-type, as this hypothesis would predict. Thus, it is possible that this mutation affects interaction with targets that have not yet been tested. Alternatively, the mutation might affect Ca²⁺ levels within neurons by interfering with the ability of CaM to buffer Ca², although such a role for CaM has not been reported.

The effects of constructs or mutations inhibiting CaMAC or CaMKII activity on the structure and function of the Drosophila neuromuscular junction are consistent with the in vitro data described above, which suggest that the Cam^3c1-encoded CaM is capable of activating these enzymes. Mutations in rutabaga (rut), which encodes CaMAC, reduce facilitation and post-tetanic potentiation at the larval neuromuscular junction (ZHONG and WU 1991 ), whereas activation of an inhibitory domain of CaMKII confers a number of behavioral, electrophysiological, and anatomical defects, including defects in courtship conditioning, an abnormal spontaneous firing of motor axon action potentials, increases in axon branching and transmitter release, and a reduction in facilitation and augmentation (GRIFFITH et al. 1993 , GRIFFITH et al. 1994 ; WANG et al. 1994 ). We have found that Cam^3c1/Cam^null larvae do not show these phenotypes; e.g., Cam^3c1/Cam^null larvae possess a normal number of axon branches and display normal ejc amplitude (in the absence of quinidine) and normal paired pulse facilitation (L. ARREDONDO, unpublished observations). Similarly, whereas application of quinidine to Cam^3c1/Cam^null substantially increases evoked transmitter release, quinidine application has little or no effect on CaMKII-inhibited larvae (L. ARREDONDO, unpublished data). Thus, it is unlikely that the Cam^3c1 mutation is exerting its effects via either CaMAC or CaMKII, but rather via an alternative target. One alternative target, the Drosophila CaM-activated protein kinase Caki, is expressed in the central nervous system, and mutants defective in caki exhibit behavioral defects related to those described here (MARTIN and OLLO 1996 ).

Other ultimate targets that could be affected in Cam^3c1/Cam^null larvae include ion channels, synaptic vesicle proteins, or other proteins located at the nerve terminal. These targets could be regulated by direct CaM binding, or indirectly via a CaM-activated regulatory enzyme. For example, it is possible that defects in synaptic vesicles that lead to increased docking or fusion of synaptic vesicles after Ca²⁺ influx could cause the observed increase in neurotransmitter release. Several proteins implicated in synaptic vesicle tethering, fusion, or recycling are regulated by CaM, including synapsins, dynamin, and rabphilin 3A (SILVA et al. 1992 ; GREENGARD et al. 1993 ; NICHOLS et al. 1994 ; FYKSE et al. 1995 ; GOOLD et al. 1995 ; LI et al. 1995 ; ROSAHL et al. 1995 ). In these cases, however, regulation is thought to occur via CaM activation of CaMKII or calcineurin, which, we have speculated, occurs normally in Cam^3c1/Cam^null.

Alternatively, defects in ion currents leading to increased Ca²⁺ influx could account for the quinidine-dependent increase in transmitter release in Cam^3c1/Cam^null larvae. In Paramecium, CaM regulates Na⁺ and K⁺ channels to alter membrane excitability and swimming behavior (reviewed by SAIMI et al. 1994 ). Such an effect could explain the defects in transmitter release and behavior that we observe in Cam^3c1/Cam^null. In addition, CaM binding to the ryanodine receptor Ca²⁺ channel in brain and muscle inhibits channel activity (GUERRINI et al. 1995 ). Defects in such an inhibition in Cam^3c1/Cam^null nerve terminals could lead to increased nerve terminal Ca²⁺ levels and the increased transmitter that we observe. Finally, ion channels are subject to regulation by phosphorylation, sometimes mediated by CaMKII, the CaM-activated phosphatase calcineurin, and the cAMP-dependent protein kinase (PKA). Although we have argued that calcineurin and CaMKII are activated normally in Cam^3c1/Cam^null larvae and adults, defective phosphorylation or dephosphorylation of ion channel subunits mediated by distinct kinases or phosphatases could also lead to the observed phenotypes.

Although our analysis to this point has not enabled an identification of the CaM target affected by the Cam^3c1 mutation, our observation that the Cam^3c1 mutation confers its phenotypes only when hemizygous suggests that two doses, but not one, of the Cam^3c1 mutant protein is sufficient to regulate this target. Thus, the Cam^3c1 protein might retain some partial ability to regulate this target.

The observation that the increases in transmitter release in Cam^3c1/Cam^null larvae are observed only in the presence of the K⁺ channel-blocking drug quinidine is consistent with the idea that a defect in ion channel underlies this aberrant transmitter release. Quinidine enhances neuronal excitability and transmitter release in mutants defective in any of several genes that encode or are thought to regulate ion channels. These genes include Shaker and Hyperkinetic, which encode K⁺ channel subunits (TEMPEL et al. 1987 ; WU et al. 1989 ; STERN and GANETZKY 1992 ; CHOUINARD et al. 1995 ), kinesin heavy chain, which encodes the motor protein kinesin that has been proposed to transport Na⁺ and K⁺ channels to nerve terminals (HURD et al. 1996 ; L. ARREDONDO, T. HILLMAN and M. STERN, unpublished results), and inebriated, which encodes a neurotransmitter transporter that is thought to affect ion channels (STERN and GANETZKY 1992 ; SOEHNGE et al. 1996 ). Thus, one might imagine that impaired activation of a K⁺ channel in Cam^3c1/Cam^null might have no observable phenotypic consequences under otherwise normal conditions; however, in combination with quinidine, which could block a distinct, functionally redundant K⁺ channel, this effect could lead to increased nerve terminal depolarization and increased Ca²⁺ influx into the nerve terminal. Our observation that the effects of Cam^3c1/Cam^null occur only at low external [Ca²⁺] is consistent with this view: the effects of most excitability mutations, including Shaker, Hyperkinetic, and inebriated, are also revealed only at low external [Ca²⁺]. This observation was proposed to result from the activation of a Ca²⁺-activated K⁺ current at higher external [Ca²⁺] (JAN et al. 1977 ). Alternatively, Cam^3c1/Cam^null could be defective in Ca²⁺ buffering. In this view, an action potential broadening conferred by quinidine application combined with reduced Ca²⁺ buffering as a consequence of Cam^3c1/Cam^null could lead to increased transmitter release.

Although mutations in several neuronal signaling genes confer defects in the pattern of motor neuron innervation of the target muscle, the morphological defects of Cam^3c1/Cam^null at the nerve terminal differ from any that have been reported previously and do not appear to result from the same mechanisms as the "activity-dependent" increases in synaptic bouton number or axonal branching (BUDNIK et al. 1990 ; ZHONG et al. 1992 ). Rather, the phenotype we have observed appears to result from defects in the formation of distinct boutons at the proper locations along the muscle surface. It is unclear why this defect in bouton formation is observed only in muscle 13. Proper synapse formation on muscle 13 may be more sensitive to altered CaM function, or perhaps distinct mechanisms control bouton formation in muscle 13 vs. other muscles. Analysis of the genes required for proper bouton formation is less characterized than for axon pathfinding or growth cone guidance [although see KESHISHIAN et al. 1996 for a recent review]. Our results suggest a role for CaM in this process.

From the nature of the phenotypic defects exhibited by Cam^3c1/Cam^null, it is possible that the Cam^3c1 mutation disrupts only one or a few cellular processes mediated by CaM. Determination of the nature of these defects will most likely require investigation, through biochemical methods, into the interactions of this mutant protein with potential target molecules.

	FOOTNOTES

¹ Present address: Department of Neurology-Box 8111, Washington University School of Medicine, 660 Euclid, St. Louis, MO 63110.
² Present address: Lark Technologies, 9545 Katy Fwy., Suite 465, Houston, TX 77024-9870.

	ACKNOWLEDGMENTS

We are grateful to PENN WHITLEY for technical assistance. This work was supported by grant 003604-028 from the Texas Advanced Technology Program to K.B. and M.S., and grants from the National Institutes of Health (GM49155) and the Robert A. Welch Foundation of Texas (C-1119) to K.B.

Manuscript received January 23, 1998; Accepted for publication June 8, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ABRAMS, T. W., K. A. KARL, and E. R. KANDEL, 1991  Biochemical studies of stimulus convergence during classical conditioning in aplysia—dual regulation of adenylate cyclase by Ca²⁺/calmodulin and transmitter. J. Neurosci. 11:2655-2665[Abstract].

BUDNIK, V., Y. ZHONG, and C.-F. WU, 1990  Morphological plasticity of motor axons in Drosophila mutants with altered excitability. J. Neurosci. 10:3754-3768[Abstract].

BUTLER, L. S., A. J. SILVA, A. ABELIOVICH, Y. WATANABE, and S. TONEGAWA et al., 1995  Limbic epilepsy in transgenic mice carrying a Ca²⁺/calmodulin-dependent kinase II subunit mutation. Proc. Natl. Acad. Sci. USA 92:6852-6855[Abstract/Free Full Text].

CHEN, C., D. G. RAINNIE, R. W. GREENE, and S. TONEGAWA, 1994  Abnormal fear response and aggressive behavior in mutant mice deficient for -calcium/calmodulin kinase II. Science 266:291-294[Abstract/Free Full Text].

CHOUINARD, S. W., G. F. WILSON, A. K. SCHLIMGEN, and B. GANETZKY, 1995  A potassium channel ß subunit related to the aldo-keto reductase superfamily is encoded by the Drosophila hyperkinetic locus. Proc. Natl. Acad. Sci. USA 92:6763-6767[Abstract/Free Full Text].

COHEN, P., and C. B. KLEE, 1988 Calmodulin: Molecular Aspects of Cellular Regulation, Vol. 5. Elsevier Press, Amsterdam.

DUDAI, Y., S. ZVI, and S. SEGEL, 1984  A defective conditioned behavior and a defective adenylate cyclase in the Drosophila mutant rutabaga. J. Comp. Physiol. 155:569-577.

FYKSE, E. M., C. LI, and T. C. SUDHOF, 1995  Phosphorylation of rabphilin-3A by Ca²⁺/calmodulin- and cAMP-dependent protein kinases in vitro. J. Neurosci. 15:2385-2395[Abstract].

GANETZKY, B. and C.-F. WU, 1982  Drosophila mutants with opposing effects on nerve excitability: genetic and spatial interactions in repetitive firing. J. Neurophys. 47:501-514[Free Full Text].

GAO, Z. H., J. KREBS, M. F. A. VAN BERKUM, W. J. TANG, and J. F. MAUNE et al., 1993  Activation of four enzymes by two series of calmodulin mutants with point mutations in individual Ca²⁺ binding sites. J. Biol. Chem. 268:20096-20104[Abstract/Free Full Text].

GHO, M., K. MCDONALD, B. GANETZKY, and W. M. SAXTON, 1992  Effects of kinesin mutations on neuronal functions. Science 258:313-316[Abstract/Free Full Text].

GOOLD, R., K. M. CHAN, and A. J. BAINES, 1995  Coordinated regulation of synapsin I interaction with F-actin by Ca²⁺ calmodulin and phosphorylation: inhibition of actin binding and bundling. Biochemistry 34:1912-1920[Medline].

GREENGARD, P., F. VALTORTA, A. J. CZERNIK, and F. BENFENATI, 1993  Synaptic vesicle phosphoproteins and regulation of synaptic function. Science 259:780-785[Abstract/Free Full Text].

GRIFFITH, L. C., L. M. VERSELIS, K. M. AITKEN, C. P. KYRIACOU, and W. DANHO et al., 1993  Inhibition of calcium/calmodulin-dependent protein kinase in Drosophila disrupts behavioral plasticity. Neuron 10:501-509[Medline].

GRIFFITH, L. C., J. WANG, Y. ZHONG, C.-F. WU, and R. J. GREENSPAN, 1994  Calcium/calmodulin-dependent protein kinase II and potassium channel subunit Eag similarly affect plasticity in Drosophila. Proc. Natl. Acad. Sci. USA 91:10044-10048[Abstract/Free Full Text].

GUERRINI, R., P. MENEGAZZI, R. ANACARDIO, M. MARASTONI, and R. TOMATIS et al., 1995  Calmodulin binding sites of the skeletal, cardiac, and brain ryanodine receptor Ca²⁺ channels: modulation by the catalytic subunit of cAMP-dependent protein kinase? Biochemistry 34:5120-5129[Medline].

GUPTAROY, B., K. BECKINGHAM, and L. C. GRIFFITH, 1996  Functional diversity of alternatively spliced isoforms of Drosophila Ca²⁺/calmodulin-dependent protein kinase II. J. Biol. Chem. 271:19846-19851[Abstract/Free Full Text].

HANSON, P. L. and H. SCHULMAN, 1992  Neuronal calcium/calmodulin-dependent protein kinases. Annu. Rev. Biochem. 61:559-601[Medline].

HEIMAN, R. G., R. C. ATKINSON, B. F. ANDRUSS, C. BOLDUC, and G. E. KOVALICK et al., 1996  Spontaneous avoidance behavior in Drosophila null for calmodulin expression. Proc. Natl. Acad. Sci. USA 93:2420-2425[Abstract/Free Full Text].

HURD, D. D., M. STERN, and W. M. SAXTON, 1996  Mutation of the axonal transport motor kinesin enhances paralytic and suppresses Shaker in Drosophila. Genetics 142:195-204[Abstract].

JAN, L. Y. and Y. N. JAN, 1976  Properties of the larval neuromuscular junction in Drosophila melanogaster. J. Physiol. 262:189-214[Abstract/Free Full Text].

JAN, Y. N., L. Y. JAN, and M. J. DENNIS, 1977  Two mutations of synaptic transmission in Drosophila. Proc. R. Soc. Lond. Ser. B 198:87-108[Medline].

KESHISHIAN, H., K. BROADIE, A. CHIBA, and M. BATE, 1996  The Drosophila neuromuscular junction: a model system for studying synaptic development and function. Annu. Rev. Neurosci. 19:545-575[Medline].

KLEE, C. B., 1991  Concerted regulation of protein phosphorylation and dephosphorylation by calmodulin. Neurochem. Res. 16:1059-1065[Medline].

KLEE, C. B. and T. C. VANAMAN, 1982  Calmodulin. Adv. Protein Chem. 35:213-321[Medline].

LI, L., L. S. CHIN, O. SHUPLIAKOV, L. BRODIN, and T. S. SIHRA et al., 1995  Impairment of synaptic vesicle clustering and of synaptic transmission, and increased seizure propensity, in synapsin I-deficient mice. Proc. Natl. Acad. Sci. USA 92:9235-9239[Abstract/Free Full Text].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, New York.

LITTLETON, J. T., M. STERN, K. L. SCHULZE, M. PERIN, and H. J. BELLEN, 1993  Mutational analysis of Drosophila synaptotagmin demonstrates its essential role in Ca²⁺-activated neurotransmitter release. Cell 74:1125-1134[Medline].

MARTIN, J. R. and R. OLLO, 1996  A new Drosophila Ca2+/calmodulin-dependent protein kinase (Caki) is localized in the central nervous system and implicated in walking speed. EMBO J. 15:1865-1876[Medline].

MARTIN, S. R., J. F. MAUNE, K. BECKINGHAM, and P. M. BAYLEY, 1992  Stopped flow studies of calcium dissociation from calcium-binding site mutants of Drosophila melanogaster. Eur. J. Biochem. 205:1107-1114[Medline].

MAUNE, J. F., 1991 Calcium binding site mutations of Drosophila melanogaster calmodulin. Ph.D. Thesis, Rice University, Houston.

MAUNE, J. F., C. B. KLEE, and K. BECKINGHAM, 1992a  Ca²⁺ binding and conformational change in two series of point mutations to the individual Ca²⁺-binding sites of calmodulin. J. Biol. Chem. 267:5286-5295[Abstract/Free Full Text].

MAUNE, J. F., K. BECKINGHAM, S. R. MARTIN, and P. M. BAYLEY, 1992b  Circular dichroism studies on calcium binding to two series of calcium binding site mutants of calmodulin. Biochemistry. 31:7779-7786[Medline].

MUKHERJEA, P. and K. BECKINGHAM, 1993  Calcium binding site mutants to calmodulin adopt abnormal conformations in complexes with model target peptides. Biochem. Mol. Biol. Int. 29:555-563[Medline].

NELSON, H. B., R. G. HEIMAN, C. BOLDUC, G. E. KOVALICK, and P. WHITLEY et al., 1997  Calmodulin point mutations affect Drosophila development and behavior. Genetics 147:1783-1798[Abstract].

NICHOLS, R. A., G. R. SUPLICK, and J. M. BROWN, 1994  Calcineurin-mediated protein dephosphorylation in brain nerve terminals regulates the release of glutamate. J. Biol. Chem. 269:23817-23823[Abstract/Free Full Text].

OHYA, Y. and D. BOTSTEIN, 1994a  Diverse essential functions revealed by complementing yeast calmodulin mutants. Science 263:963-966[Abstract/Free Full Text].

OHYA, Y. and D. BOTSTEIN, 1994b  Structure-based systematic isolation of conditional-lethal mutations in the single yeast calmodulin gene. Genetics 138:1041-1054[Abstract].

PRESTON, R. R., J. A. KINK, R. D. HINRICHSEN, Y. SAIMI, and C. KUNG, 1991  Calmodulin mutants and Ca²⁺-dependent channels in Paramecium. Annu. Rev. Physiol. 53:309-319[Medline].

RICHARDS, S., T. HILLMAN, and M. STERN, 1996  Mutations in the Drosophila pushover gene confer increased neuronal excitability and spontaneous synaptic vesicle fusion. Genetics 142:1215-1223[Abstract].

ROSAHL, T. W., D. SPILLANE, M. MISSLER, J. HERZ, and D. K. SELIG et al., 1995  Essential functions of synapsins I and II in synaptic vesicle regulation. Nature 375:488-493[Medline].

SAIMI, Y., K. Y. LING and C. KUNG, 1994 Calmodulin sensitive channels, pp. 435–443 in Handbook of Membrane Channels, edited by C. PERACCHIA, Academic Press, New York.

SILVA, A. J., C. F. STEVENS, S. TONEGAWA, and Y. WANG, 1992  Deficient hippocampal long-term potentiation in -calcium-calmodulin kinase II mutant mice. Science 257:201-206[Abstract/Free Full Text].

SINGH, S. and C.-F. WU, 1989  Complete separation of four potassium currents in Drosophila. Neuron 2:1325-1329[Medline].

SMITH, V. L., K. E. DOYLE, J. F. MAUNE, R. P. MUNJAAL, and K. BECKINGHAM, 1987  Structure and sequence of the Drosophila melanogaster calmodulin gene. J. Mol. Biol. 196:471-485[Medline].

SOEHNGE, H., X. HUANG, M. BECKER, P. WHITLEY, and D. CONOVER et al., 1996  A neurotransmitter transporter encoded by the Drosophila inebriated gene. Proc. Natl. Acad. Sci. USA 93:13262-13267[Abstract/Free Full Text].

SOLEM, M., T. MCMAHON, and R. O. MESSING, 1995  Depolarization-induced neurite outgrowth in PC12 cells requires permissive, low level NGF receptor stimulation and activation of calcium/calmodulin-dependent protein kinase. J. Neurosci. 15:5966-5975[Abstract].

STERN, M. and B. GANETZKY, 1992  Identification and characterization of inebriated, a gene affecting neuronal excitability in Drosophila. J. Neurogen. 8:157-172[Medline].

STRITTMATTER, S. M., C. FANKHAUSER, P. L. HUANG, H. MASHIMO, and M. C. FISHMAN, 1995  Neuronal pathfinding is abnormal in mice lacking the neuronal growth cone protein GAP-43. Cell 80:445-452[Medline].

TEMPEL, B. L., D. M. PAPAZIAN, T. L. SCHWARZ, Y. N. JAN, and L. Y. JAN, 1987  Sequence of a probable potassium channel component encoded at the Shaker locus of Drosophila. Science 237:770-775[Abstract/Free Full Text].

UMBACH, J. A., K. E. ZINSMAIER, K. K. EBERLE, E. BUCHNER, and S. BENZER et al., 1994  Presynaptic dysfunction in Drosophila csp mutants. Neuron 13:899-907[Medline].

VANBERKUM, M. F. A. and C. S. GOODMAN, 1995  Targeted disruption of Ca²⁺-calmodulin signaling in Drosophila growth cones leads to stalls in axon extension and errors in axon guidance. Neuron 14:43-56[Medline].

WANG, J., J. J. RENGER, L. C. GRIFFITH, R. J. GREENSPAN, and C. F. WU, 1994  Concomitant alterations of physiological and developmental plasticity in Drosophila CaM kinase II-inhibited synapses. Neuron 13:1373-1384[Medline].

WILLIAMS, E. J., B. MITTAL, F. S. WALSH, and P. DOHERTY, 1995  A Ca²⁺/calmodulin kinase inhibitor, KN-62, inhibits neurite outgrowth stimulated by CAMs and FGF. Mol. Cell. Neurosci. 6:69-79[Medline].

WU, C.-F., M.-C. TSAI, M.-L. CHEN, Y. ZHONG, and S. SINGH et al., 1989  Actions of dendrotoxin on K⁺ channels and neuromuscular transmission in Drosophila melanogaster, and its effects in synergy with K⁺ channel-specific drugs and mutations. J. Exp. Biol. 147:21-41[Abstract/Free Full Text].

ZHONG, Y. and C.-F. WU, 1991  Altered synaptic plasticity in Drosophila memory mutants with a defective cyclic AMP cascade. Science 251:198-201[Abstract/Free Full Text].

ZHONG, Y., V. BUDNIK, and C.-F. WU, 1992  Synaptic plasticity in Drosophila memory and hyperexcitable mutants: role of cAMP cascade. J. Neurosci. 12:644-651[Abstract].

This article has been cited by other articles:

				A. S. Rawls, S. A. Schultz, R. D. Mitra, and T. Wolff Bedraggled, a Putative Transporter, Influences the Tissue Polarity Complex During the R3/R4 Fate Decision in the Drosophila Eye Genetics, September 1, 2007; 177(1): 313 - 328. [Abstract] [Full Text] [PDF]

				B. Wang, K. M. C. Sullivan, and K. Beckingham Drosophila Calmodulin Mutants With Specific Defects in the Musculature or in the Nervous System Genetics, November 1, 2003; 165(3): 1255 - 1268. [Abstract] [Full Text] [PDF]

				J. E. Keen, R. Khawaled, D. L. Farrens, T. Neelands, A. Rivard, C. T. Bond, A. Janowsky, B. Fakler, J. P. Adelman, and J. Maylie Domains Responsible for Constitutive and Ca2+-Dependent Interactions between Calmodulin and Small Conductance Ca2+-Activated Potassium Channels J. Neurosci., October 15, 1999; 19(20): 8830 - 8838. [Abstract] [Full Text] [PDF]

				E. Van Uden, Y. Sagara, J. Van Uden, R. Orlando, M. Mallory, E. Rockenstein, and E. Masliah A Protective Role of the Low Density Lipoprotein Receptor-related Protein against Amyloid beta -Protein Toxicity J. Biol. Chem., September 22, 2000; 275(39): 30525 - 30530. [Abstract] [Full Text] [PDF]