Caenorhabditis elegans beta-G spectrin is dispensable for establishment of epithelial polarity, but essential for muscular and neuronal function

Abstract

The Caenorhabditis elegans genome encodes one alpha spectrin subunit, a beta spectrin subunit (beta-G), and a beta-H spectrin subunit. Our experiments show that the phenotype resulting from the loss of the C. elegans alpha spectrin is reproduced by tandem depletion of both beta-G and beta-H spectrins. We propose that alpha spectrin combines with the beta-G and beta-H subunits to form alpha/beta-G and alpha/beta-H heteromers that perform the entire repertoire of spectrin function in the nematode. The expression patterns of nematode beta-G spectrin and vertebrate beta spectrins exhibit three striking parallels including: (1) beta spectrins are associated with the sites of cell-cell contact in epithelial tissues; (2) the highest levels of beta-G spectrin occur in the nervous system; and (3) beta spectrin-G in striated muscle is associated with points of attachment of the myofilament apparatus to adjacent cells. Nematode beta-G spectrin associates with plasma membranes at sites of cell-cell contact, beginning at the two-cell stage, and with a dramatic increase in intensity after gastrulation when most cell proliferation has been completed. Strikingly, depletion of nematode beta-G spectrin by RNA-mediated interference to undetectable levels does not affect the establishment of structural and functional polarity in epidermis and intestine. Contrary to recent speculation, beta-G spectrin is not associated with internal membranes and depletion of beta-G spectrin was not associated with any detectable defects in secretion. Instead beta-G spectrin-deficient nematodes arrest as early larvae with progressive defects in the musculature and nervous system. Therefore, C. elegans beta-G spectrin is required for normal muscle and neuron function, but is dispensable for embryonic elongation and establishment of early epithelial polarity. We hypothesize that heteromeric spectrin evolved in metazoans in response to the needs of cells in the context of mechanically integrated tissues that can withstand the rigors imposed by an active organism.

Figure 1

Cloning of C. elegans β-G spectrin cDNAs, including two alternatively spliced variants. (A) A partial cDNA clone, CEESK92, was used to screen a C. elegans mixed stage λZAP cDNA library. Positions of representative partial cDNAs (SCSM-II and SCSM-2) and the reverse transcriptase–PCR sequence corresponding to the β-G spectrin (BGS-1) mRNA are shown. (B) Bgs-1 exon distribution. The transcriptional unit is distributed over 13 exons, of which at least two are alternately spliced (asterisks). The BGS-1–predicted ORFs code for two polypeptides of 2,302 and 2,257 amino acids (see below). The coding sequence is completely contained within the unsequenced cosmid T23C12, and distributed between sequenced cosmids K11C4 and T19F4. (C) An alternate splice. The two exons marked by asterisks in B are alternatively spliced in a small number of transcripts to introduce 45 novel amino acids before the PH domain. Only the region flanking the alternate splice (extending into the PH domain) was included in the alignment. (D) Genomic location of C. elegans spectrin genes. Sequencing of the genome has identified three candidate spectrin genes: an α subunit (spc-1), a β-G subunit (bgs-1/unc-70), and a β-H (heavy) subunit (sma-1). The chromosome number and genes mapped to the vicinity of each spectrin subunit are indicated. Bars: (A and B) 1 kb; (D) 0.5 kb.

Figure 2

Three spectrin genes of C. elegans. (A) Domain organization of α and β spectrin polypeptides. C. elegans orthologues are indicated. Spectrin is an (α-β)₂ tetramer; an α-β-G tetramer is shown. (B) Evolutionary relationship of β spectrin subunits. A UPGMA tree showing relative homology between five β-spectrin subunits from different species was plotted using Geneworks v2.5.1 software. Of the three identified human β-spectrin proteins (SPTB, SPTBN1, and SPTBN2), the two nonerythroid β-spectrins SPTBN1 and SPTBN2 are more closely related (65% identity). The two invertebrate β-G spectrins show ∼58% identity, whereas ∼32% of β-spectrin residues are conserved among all five proteins. (C) Dot matrix alignments of C. elegans predicted spectrin polypeptides with their human (α and β-G) or Drosophila (β_H) orthologues. Note the additional diagonals that arise from the multiple spectrin repeats that make up the bulk of each subunit. (D) Comparison of PH domains of β_G-spectrins from Homo sapiens (SPTBN1), Drosophila (DROBSPEC), and C. elegans (BGS-1, shorter spliceform). Residues used in the alignment are indicated. Note that conserved residues fall in loop regions that determine the substrate specificity of PH domains.

Figure 3

β-G spectrin is associated with the plasma membrane at sites of cell–cell contact throughout development in C. elegans. (A) Immunoblot. Protein lysates were prepared from different developmental stages of N2 wild-type worms, and a Western blot was performed using affinity-purified rabbit polyclonal β-G spectrin antibody (see Materials and Methods). A single band of M _r = 225–235 kD is seen at all stages. D = Dauer stage, E = embryos, L = L2-L3 larvae, and M = mixed stages. (B) β-G-spectrin levels during embryogenesis. All embryos in the uterus of a single adult hermaphrodite were stained by indirect immunofluorescence using β-G spectrin antibody and an FITC-coupled secondary antibody (see Materials and Methods). Levels of β-G spectrin are low, but are detectable in early embryos (arrowhead indicates four-cell stage), increase after the ∼100-cell stage (short arrow), and continue to dramatically increase throughout embryogenesis. The image in B has been color coded using a heat scale (Adobe Photoshop) to indicate the sharp increase in levels of β-G spectrin after gastrulation and in several tissues such as the developing nervous system in older embryos. The heat scale ranges from 0 (black) to 255 (white) to indicate the lowest and highest amounts of spectrin, respectively. (C). β-G spectrin localization during embryogenesis. β-G spectrin is associated with plasma membranes at sites of cell–cell contact (i and ii, arrows), starting at the two-cell stage, with cell membranes in the gut (iii, arrows; v, gt), in epidermal cells (iv, arrow indicates row of lateral seam cells), and in the developing pharynx (v, ph). Levels are particularly high in the nervous system, with intense labeling at the tips of the sensillae (vi and vii, arrowhead), and in the developing nerve ring (vii and viii, arrow) and nerve cords (see E below). (D) β-G spectrin in the adult. A confocal optical section centered at the level of epidermis in an adult hermaphrodite stained for β-G spectrin. Shows labeling of the lateral domains of seam cells (s), the cell borders of the spindle-shaped body-wall muscle cells (m), and the bundles of commissural axons (ca). (E) An optical section taken through the same worm, 25 μm below D. The intestine (gut) continues to show staining, but the highest levels of β-G spectrin are associated with the pharyngeal region (ph indicates the metacorpus), the spermathecae (sp), the tips of the sensillae (sn), the nerve ring (nr), and ventral nerve cord (vnc). Bars: (B) 25 μm; (C) 10 μm; (D and E) 50 μm.

Figure 4

Depletion of β-G spectrin early in embryogenesis by RNAi. A noninjected adult hermaphrodite (A) and an adult (B) injected with β-G spectrin dsRNA were killed, and their embryos were processed for immunofluorescence labeling (see Materials and Methods). Double labeling with β-G spectrin antibody (green) and DNA (Hoechst, red) revealed that β-G spectrin is depleted in the F1 progeny of hermaphrodites injected with bgs-1 dsRNA. Staining is reduced even in very early embryos (arrowhead indicates two-cell embryos in each panel), and the protein levels are demonstrably lower compared with similarly staged wild-type animals all through embryogenesis. β-G spectrin in the somatic spermatheca (arrows) serves as an internal staining control. Note that some of the intense FITC labeling of the spermatheca has bled into the Hoechst channel. (C) Measuring the effectiveness of RNAi in depleting animals of β-G spectrin. The amounts of β-G spectrin in the control, wild-type, and RNAi progeny, which are shown in an immunoblot, were quantitated using a PhosphorImager (ImageQuaNT). Taking the background signal into account, the counts were divided by the number of animals used in each lysate (250) to give an arbitrary unit of β-G spectrin per animal as shown in the graph. The numbers indicate that RNAi is effective in depleting β-G spectrin by at least 98%. Bars, 25 microns.

Figure 5

Phenotypic effects of α, β-G, and β-H spectrin depletion by RNAi. dsRNA corresponding to the three C. elegans spectrin genes (spc-1, bgs-1, and sma-1 were microinjected into the syncytial gonad of young adults (see Materials and Methods). Progeny born 8 h after injection were analyzed and their phenotypes were recorded. bgs-1(RNAi) animals arrest as early uncoordinated L1 larvae (B), sma-1(RNAi) animals survive to form small (sma) viable adults (C), and both spc-1(RNAi) and [bgs-1(RNAi);sma-1(RNAi)] double mutants arrest as early small uncoordinated larvae (D and E). For each spectrin RNAi experiment, differential interference micrographs and indirect immunofluorescence labeling using bgs-1 antibody (green) are shown. The mAb MH27 (red) recognizes adherens junctions, and was used as a permeabilization control in double labeling experiments for bgs-1, sma-1, and [(bgs-1;sma-1) RNAi larvae. Note that the low levels of residual β-G spectrin appear to stain the nerve rings in G and J. Bars, 10 μm.

Figure 7

Depletion of each of the three C. elegans spectrins does not alter localization of proteins to distinct domains in the epithelia. (A) Formation of apicolateral adherens junctions. Control (left column) and bgs-1(RNAi) (middle) embryos at the bean stage were double labeled for the antibody against β-G spectrin (green) and the mAb MH27 (red), which recognizes an adherens junction component. Depletion of β-G spectrin (middle) did not affect localization of MH27 to the adherens junctions at the apicolateral domain of the epidermal epithelia (arrowheads) or the intestinal cells (arrows). RNAi using both bgs-1 and sma-1 β spectrin dsRNAs still did not prevent formation of properly localized adherens junctions in the gut (arrows, right column). (B) Localization of lateral domain markers. bgs-1(RNAi) animals at the bean stage were stained for the β-G spectrin-binding protein ankyrin/UNC-44 (i) and the ankyrin binding L1-related cell adhesion molecule LAD-1 (ii). Both localized normally to the lateral membranes of epidermal (arrowheads) and gut cells (arrows). (C) spc-1(RNAi) larvae were stained with the BGS-1 antibody. Though they arrest at the L1 stage with morphogenetic defects, these animals continued to express normal levels of β-G spectrin limited to the lateral membranes of gut cells. Inset shows a higher magnification of the boxed region, arrows point to the lateral domains of the gut cells. (D) In N2 adults (I), β-G spectrin clearly stains only the lateral but not the apical domains of the gut cells. In sma-1(ru18)–null mutants (ii), the absence of β-H spectrin does not affect the retention of β-G spectrin at the lateral membranes of the gut (arrows). Bars: (A–C) 10 μm; (D) 50 μm.

Figure 8

RNAi depletion of β-G spectrin does not affect the ultrastructural morphology of L1 larval intestinal epithelia. bgs-1(RNAi) and control L1 larvae were processed for electron microscopy (see Materials and Methods). Transverse sections, which were taken through the middle of the animal, show that gross morphology appears relatively normal in recently arrested L1 larvae. The seam cells polarize to deposit the distinctive cuticle of the alae (al). The adherence junction (aj) between the two intestinal cells shows appropriate localization and structure, and the intestinal microvilli (mv) and yolk granules (y) are abundant and morphologically normal. In particular, the lateral cell membranes (lm) separating the two cells appear distinct and unaltered at this stage. Note, coelomocyte (Cc) in pseudocoelom of B.

Figure 9

β-G spectrin is associated with multiple structures of the sarcolemma of body-wall muscle cells. (A) Optical sections in two focal planes through the same body-wall muscle cell stained with β-G spectrin antibody. L is centered closer to the myofilament lattice, whereas S is centered near the sarcolemma and closer to the overlaying epidermal cell. (B) Double labeling to visualize β-G spectrin and β-integrin (MH25). The two proteins colocalize at the base of dense bodies (arrowheads) and at the base of M-lines (arrow). β-G spectrin staining extends beyond the dense bodies and appears to lie close to the sarcolemma adjoining the I-bands. (C) Double labeling to visualize β-G spectrin and the myosin heavy chain A (MHC-A), the latter is restricted to the center of A-bands. The broader bands of spectrin staining (arrowheads) alternate with the MHC-A staining (arrows), while there is a faint staining of BGS-1 in the region corresponding to the M-line (vertical arrows). (D) Double labeling with antibodies against β-G spectrin and the hemidesmosomal component recognized by the mAb MH5 reveals that the patterns are very similar, but the two proteins do not colocalize. The animal was fixed with methanol, which reduces staining along the longitudinal tracts. All images are to the same scale. Bars, 10 μm.

Figure 10

Depletion of β-G spectrin leads to progressive perturbation of the myofilament lattice of body-wall muscle. (A) Thin filaments. L1 larvae were fixed and stained with biotinylated phalloidin, followed by Cy5-coupled streptavidin to visualize F-actin in body-wall muscle. The F-actin–containing thin filaments in bgs-1(RNAi) L1 larvae (ii) are disorganized compared with controls (i). Faint staining from the pharynx is seen in both panels. Animals were costained with β-G spectrin antibody to ensure its depletion by RNAi (not shown). (B) Thick filaments. Labeling with the mAb DM5.6 that recognizes the myosin heavy chain A (mhcA). bgs-1(RNAi) larvae (ii) present an unraveled appearance of the A-bands. Bars, 5 μm.

Figure 11

Depletion of β-G spectrin from the neurons of C. elegans larvae leads to defects in cell body size, location, and in neurite morphology. The dye-filling amphid neurons of (A) N2 and (B) bgs-1(RNAi) larvae were visualized using the cationic membrane tracer dye DiI (see Materials and Methods). The dye was visualized on the rhodamine channel (A ii and B ii). bgs-1(RNAi) larvae displayed a variety of defects, including cell bodies that are enlarged (large arrow) or ectopic (small arrow). Vacuoles (B, arrowhead in i and ii) were often observed in neurites and adjacent sheath cells, and resembled those formed by cell death in sick strains. The axonal processes leading into the nerve ring (black arrowhead) and the dendritic tree (black arrow) are marked. Bars, 5 μm.