Unc gene products compose various parts of the body wall muscle (Fig. 1). We identified 45 H. glycines est homologs of C. elegans unc genes (Hg-unc) (Figs. 2 and 3). We confirmed the identification of the Hg-unc genes by performing manual blast searches of the C. elegans unc genes in Genbank. We also identified other H. glycines ests (dystrophin [Hg-dys-1], neprilysin [Hg-nep-1], actin [Hg-act-1], talin [Hg-talin], pat-6 [Hg-pat-6]) whose mutants exhibit unc phenotypes or whose protein products interact with UNC proteins in C. elegans. However, the original unc mutant screens did not identify them (Figs. 2 and 3).

We identified a decline in transcript abundance for Hg-unc-87 during the transition from the mobile to the sedentary phase of the H. glycines lifecycle 19. This observation indicates that microfilament degradation occurs during muscle wasting. Bioinformatics analyses identified several H. glycines ests that are homologous to C. elegans thin filament genes, including Hg-act-1, Hg-unc-27, Hg-unc-60A, Hg-unc-60B and Hg-unc-78 (Figs. 2 and 3). RT-PCR revealed a substantial decline in actin transcript abundance occurring between the J2 stage and 15 dpi nematodes (Fig. 4). RT-PCR revealed a substantial decline in transcript abundance of Hg-unc-27 occurring between the J2 stage and 15 dpi nematodes (Fig. 4). We examined the expression profile of the two Hg-unc-60 isoforms (A and B) and Hg-unc-78. Hg-unc-60A is the non-muscle unc-60 isoform, while Hg-unc-60B is the muscle-specific isoform) RT-PCR of Hg-unc-60A reveals little change in transcript abundance occurring throughout the H. glycines lifecycle (Fig. 4). However, RT-PCR reveals a substantial decline in transcript abundance occurring for Hg-unc-60B between the J2 stage and 15 dpi nematodes (Fig. 4). The decline in transcript abundance for Hg-unc-60B and not Hg-unc-60A, occurring between the J2 stage and 15 dpi nematodes, is in agreement with its muscle-specific activity. RT-PCR reveals little change in transcript abundance occurring throughout the H. glycines lifecycle for Hg-unc-78 (Fig. 4).

Bioinformatics analyses identified H. glycines ests homologous to C. elegans thick filament genes (Figs. 2 and 3). Transcript abundance of H. glycines unc genes whose homologous gene products compose thick filaments in C. elegans was measured using RT-PCR. RT-PCR revealed a substantial decline in transcript abundance of Hg-unc-15 and Hg-unc-54 occurring between the J2 stage and 15 dpi nematodes (Fig. 5). Furthermore, transcript levels of Hg-unc-89 also decline between the J2 stage and 15 dpi nematodes (Fig. 5).

Bioinformatics analyses also identified H. glycines ests homologous to C. elegans focal adhesion genes (Figs. 2 and 3). Hg-unc-97 transcript levels decrease in abundance between the J2 and 15 dpi nematodes, as shown by RT-PCR (Fig. 6). We explored the focal adhesion complex further by examining the transcript abundance of Hg-unc-112, Hg-pat-6 and Hg-talin. Hg-unc-112, Hg-pat-6 and Hg-talin transcript levels decrease between the J2 stage and 15 dpi nematodes (Fig. 6). Bioinformatics analyses did not identify homologs of other focal adhesion complex proteins.

Bioinformatics analyses identified H. glycines ests homologous to C. elegans genes whose protein products function in other aspects of muscle biology (Figs. 2 and 3) These H. glycines unc genes include Hg-unc-9, Hg-unc-22 (twitchin), Hg-unc-31(CAPS), Hg-unc-52 (perlecan), Hg-unc-101, Hg-unc-115, Hg-unc-110 (Hg-twk-18), Hg-dys-1, and Hg-nep-1. RT-PCR analysis indicated that modest changes in transcript abundance occur for Hg-unc-9, Hg-unc-22, Hg-unc-31, Hg-unc-52, Hg-unc-101, Hg-unc-115, Hg-dys-1, and Hg-nep-1 between the J2 stage and 15 dpi nematodes (Fig. 7). RT-PCR also indicated that Hg-unc-110 transcript abundance decreases substantially between the J2 stage and 15 dpi nematodes (Fig. 7).

Body wall muscle degradation accompanies the sedentary phase of H. glycines as it feeds from the syncytium. Thus, important transcriptional, translational, and post-translational changes occur at this time. We began our analysis of H. glycines muscle wasting by identifying H. glycines homologs of C. elegans muscle genes. We then identified the transcript abundance of those genes whose protein products compose the acto-myosin complex, muscle focal adhesion complex, neuromuscular connections and potassium channels.

The acto-myosin complex is composed of interdigitating thin and thick filaments that are bundled by UNC-87 293031. Previously, we observed a decline in transcript abundance of Hg-unc-87 19. This demonstrated that depletion of components of the acto-myosin complex may occur during the sedentary phase of the H. glycines lifecycle. Actin and troponin I (unc-27) are primary components of the thin filaments. Actin is not classified as an unc gene. However, the unc-92 mutant of C. elegans maps to the actin locus and may actually be actin. Recently, Willis et al. 11, found that the actin family in C. elegans is composed of five highly conserved isoforms (act-1–5) and yields an unc phenotype 11. Only one actin gene is present in H. glycines 32. Unc-27 is involved in thin filament maintenance. UNC-27 forms a complex with troponin C (PAT-10) and troponin T 333435 to accomplish calcium-dependent regulation of the acto-myosin interaction 36. Mutant unc-27 disorganizes dense body positioning. Mutant unc-27 causes less well-defined sarcomeres with small regions of thin filaments interspersing within the overlap of A-bands 37. We found that, as expected, Hg-act-1 and Hg-unc-27 experience a substantial decrease in expression between J2 and 15 dpi nematodes

The dynamic nature of actin filaments is under control of the actin interacting proteins UNC-60 and UNC-78. UNC-60 is the

a

d

f

a

i

p

Thick filaments are major components of muscles. In C. elegans, myosin (UNC-54), paromyosin (UNC-15) and myosin heavy chain A (MYO3) compose thick filaments. Thick filaments are anchored to the M-line on one side and bound to the dense body on the side by the protein titin 44. UNC-89 organizes muscles by assembling thick filaments into A-bands 45. UNC-89 is also essential for M-line assembly 45. There are three UNC-89 isoforms n C. elegans 45. Our RT-PCR analysis demonstrates a decrease in transcript abundance for Hg-unc-15, Hg-unc-54 and Hg-unc-89 occurring during muscle wasting. Thus, a decrease in transcript abundance for actin and myosin gene products occurs during muscle wasting as nematodes are becoming sedentary during their parasitic feeding stages. Loss of muscle mass occurs in mutants for muscle genes. For example, loss of muscle mass is a characteristic of DMD, caused by dys-1 mutants. However, in C. elegans, DMD-like muscle defects also require dystrobrevin (DYB-1). A microarray experiment explored the complexities of the dys-1 mutant background 46. Microarrays of dys-1 revealed 44 total probe sets are induced while 71, including unc-89, are suppressed 46. It is not clear how a decrease in transcript abundance of unc-89 is involved in DMD. Differential expression of myosin transcripts was also observed in that study 46.

The focal adhesion complex is composed of numerous proteins. In C. elegans, UNC-97 is part of the PINCH family of proteins that are composed of five

l

i

m

Our bioinformatics analysis identified the focal adhesion complex genes Hg-unc-97, Hg-unc-112, Hg-talin, and Hg-pat-6. A modest decrease in transcript abundance occurs for these genes between the J2 and 15 dpi nematodes. These results demonstrate that the deterioration of focal adhesion sites, by depletion of Hg-UNC-97, Hg-UNC-112, Hg-TALIN, and Hg-PAT-6, may not be a major contributor to body wall muscle wasting.

Unc gene products perform other important roles in muscle biology. For example, UNC-9, is a neuromuscular gap junction protein 55; UNC-22 (twitchin) is involved in muscle A-band structure 5657; UNC-31(CAPS) is involved in the neuromuscular junction and neurosecretion 58, UNC-52 (perlecan), is a muscle basement membrane heparan sulfate proteoglycan protein 59. Mutant unc-52 can also exhibit a pat phenotype, depending on the mutant allele 60; UNC-101, is a clathrin-associated protein having intense expression in muscles and pharynx; UNC-115, is an actin-binding, LIM domain containing protein that is involved in neuronal axon guidance 6162 and UNC-110, is a potassium channel subunit protein involved in body wall muscle control 63. Dystrophin (dys-1) and neprilysin (nep-1) are other genes whose mutants exhibit unc-like phenotypes. The dys-1 gene is the C. elegans Duchenne muscular dystrophy homolog. The dys-1 gene product is part of the dystrophin-glycoprotein complex that is found in the plasma membranes of muscle cells. The dystrophin-glycoprotein complex is responsible for linking the intracellular cytoskeleton to the extracellular matrix and thought to be important for organizing signal molecules 64 and mechanical integrity 65. The nep-1 gene product is involved in the neuronal network of pharyngeal pumping 66.

We observed only modest changes in gene expression occurring for many of the other H. glycines unc gene homologs. However, the H. glycines homolog of the body wall muscle-specific potassium channel protein, UNC-110, experiences a substantial decrease in transcript abundance between the J2 and 15 dpi nematodes. This decrease in transcript abundance is similar to the decrease in transcript abundance shown for the acto-myosin genes. It is not clear how a decrease in abundance for potassium channel proteins like Hg-UNC-110 would contribute to the sedentary nature of H. glycines during later stages of parasitism. However, potassium channels do perform major roles in muscle function in C. elegans 636768. At least 42 genes exist in the C. elegans genome that encode

tw

K

Plant and nematode materials were grown at the United States Department of Agriculture Soybean Genomics and Improvement Laboratory as described previously 21 according to the moisture replacement system 69. Our study uses eggs and J2 stage nematodes that are composed of male and female nematodes. This was done because at this time it is not possible to distinguish between immature male and females at the egg and J2 stages. Hatching and subsequent migration of J2s are identical between male and female nematodes. Thus, it is likely that, concerning muscle biology, males and females are nearly identical at the egg and J2 stages. The later stages we use (15 and 30 dpi) are composed entirely of sedentary parasitic female nematodes because that was the focus of this study.

Briefly, G. max cv. Peking seeds were grown in a sand mix in standard greenhouse conditions. To promote hatching, eggs from the H. glycines isolate TN8 (susceptible reaction) were incubated in sterile water at room temperature on a rotary shaker at 25 rpm. After two days on the rotary shaker, the J2s were collected and concentrated by centrifugation to approximately 5,200 J2/ml. Replicate experiments were performed and completed by running one experiment to completion and then collecting data. A second experiment was then run at approximately one month later after the first experiment was completed. Thus, we isolated different isolations of H. glycines eggs, J2, 15 dpi and 30 dpi nematodes for each experiment. Four plants contained in a beaker were inoculated with 5,200 J2 nematodes. For RT-PCR experiments, recovery of the 15 and 30 dpi H. glycines samples from G. max roots was performed according to 69 and also done previously in our lab 19. Briefly, H. glycines-infected G. max roots grown for either 15 or 30 dpi were dipped into water to remove sand. At 15 and 30 dpi, female H. glycines are partially emerged from the root, facilitating collection of pure nematode samples. Roots were lightly massaged to liberate female H. glycines into a sieve of 150 μm pore size (VWR Scientific; Bridgeport, NJ). This filtration step would remove any additional male nematodes that were remaining in the root. To obtain egg and J2 samples, mature females were harvested at 30 dpi and crushed. The eggs were sifted through a 250 μm sieve and captured onto a 25 μm sieve. The eggs were hatched for two days in distilled water on a rotary shaker at 25 rpm. Pure J2 suspensions are made by sifting them through a 41 μm nylon mesh and collected with a low-speed centrifugation. The RT-PCR samples (J2, 15 dpi female, 30 dpi female or eggs) were flash-frozen in liquid nitrogen and ground to a fine powder using mortar and pestle chilled in liquid nitrogen. Total RNA extraction was performed using the method of Mujer et al. 70.

The CeHg bioinformatics database 26 is a database containing 300,773 ests and 6,630 genomic sequences from C. elegans and 24,438 ests and 231 genomic sequences from H. glycines (May, 2006). These C. elegans and H. glycines sequences were used to create a local database using SQLServer2000. The sequences were imported into our local database. Subsequently we created a unigene set using the contig assembly program Seqman (DNAStar Inc.; Madison, WI) resulting in 3,782 contigs of 2 or more sequences and 4,522 singletons for H. glycines. These sequences were then blasted against the local C. elegans database. Parsing of the results of the blast searches was done with customized Perl scripts. These scripts extracted the best hits from the blast results, E-value, score and identities values. The parsed results were imported back into the database. SQL scripts were written to query the CeHg database for C. elegans genes having high homology. Our data base was then linked to WormBase 71 and PubMed 72 to identify H. glycines ests homologous to C. elegans unc genes. These results then were confirmed by performing manual blast searches of each C. elegans unc gene against the H. glycines sequences.

RNA was extracted from nematodes as previously described and treated with DNase I to remove genomic DNA. The cDNA was reversed transcribed from RNA using SuperScript First Strand Synthesis System for RT-PCR (Invitrogen; Grand Island, NY) with oligo d(T) as the primer according to manufacturer's instructions. All the primer sets were initially tested for specificity with a mixture of RNAs for all stages of nematodes. Genomic DNA contamination was assessed by PCR as described previously 19. We performed this experiment to identify any contaminating genomic DNA that may exist in our cDNA. We used Hg-unc-87 PCR primers (forward primer: 5'GACAACACGGAGATTCCACTTCAG3'; reverse primer, 5'CTGGTCTGGTCGATGCTCTGCTC3') that amplify different size fragments in the presence of genomic DNA as compared to pure cDNA. RT-PCR reactions containing no template and reactions using RNA processed in parallel but with no Superscript reverse transcriptase also served as controls for RT-PCR and produced no amplicon. After we determined that no contaminating genomic DNA existed in our cDNA, we performed RT-PCR. Relative quantities of expression using their respective primers (Fig. 8) were determined using an Mx3000P Real-Time PCR system following manufacturer's instructions (Stratagene; La Jolla, CA). DNA accumulation was measured using SYBR Green and ROX was used as reference dye. Only one product was present in each reaction as indicated by the SYBR Green dissociation curves of amplified products and by assay of terminal reactions by gel electrophoresis in 1% TBE agarose, thus ensuring that the product was of proper size. Template DNA was denatured for 10 minutes at 96°C, followed by PCR cycling temperatures set for denaturing for 30 seconds at 96°C, annealing for 60 seconds at 55°C and extension for 30 seconds at 72°C. The standard curve for the expression comparisons was constructed from the J2 stage sample. The J2 stage sample was diluted over a five-log range and used in parallel RT-PCR assays. All RT-PCR assays were conducted in triplicate. Threshold cycle (C_t) values were plotted against the dilution series. PCR efficiencies were equal between the target and endogenous control. C_t values and relative abundance were calculated using software supplied with the Mx3000P Real-Time PCR system. Our RT-PCR data was standardized against an est (CB380016) determined to experience no change in expression during H. glycines development. The relative abundance of mRNA was compared to that of CB380016 in the different sample types to calculate fold change. For each gene, a ratio was established between the control (CB380016) and the gene of interest (GOI) for the egg J2, 15 dpi and 30 dpi samples. To calculate fold expression, the ratio between CB380016 and the GOI at 15 dpi was set to a value of one. Other fold expression values for egg, J2 and 30 dpi were calculated using the ratio obtained at 15 dpi for GOI as the denominator. The ratio of the GOI for egg, J2 and 30 dpi, respectively, was used as the numerator. The value obtained after calculation was fold expression for those time-points. Standard error was used in the analyses.