From a mutagenesis screen in the nematode C. elegans we isolated the mutant bar18, showing an accumulation of muscle cell nuclei around the posterior pharyngeal bulb of the worm. Quantification of the overall amount of body wall muscle nuclei, based on the muscle-specific reporter myo-3p::gfp::NLS, revealed that the number of nuclei in bar18 mutants is unchanged compared to WT worms. The accumulation of muscle nuclei around the posterior pharyngeal bulb is due to a positioning defect, which can be precisely quantified by subdividing the worm into head, neck, and posterior body segments.

Whole-genome sequencing revealed that bar18 animals carry a mutation in the KASH-domain gene unc-83 causing a premature STOP. An additional unc-83 mutant allele recapitulates the phenotype, as does a mutant allele of UNC-84, a SUN-domain containing protein that interacts with UNC-83. UNC-83 and UNC-84 belong to a Linker of Nucleoskeleton and Cytoskeletonnuclear (LINC) complex that bridges the nuclear lamina with the cytoskeleton. SUN and KASH domain proteins are conserved in mammals and mutations in the corresponding genes have been linked to cancer, autism, muscular dystrophy and other diseases. Additionally, LINC complexes that function in nuclear migration have also been identified in mammals.

We were able to rescue the unc-83 mutant phenotype by expressing the WT gene under a muscle-specific (myo-3p) promoter, demonstrating that the effect is cell autonomous. Mutations in either unc-83 or unc-84 have previously been linked to nuclear migration defects in P cells, intestinal cells and hyp7 hypodermal precursors but not in muscles. Whether the mis-positioning of muscle nuclei is due to migration or anchoring defects still needs to be determined.


A non-functioning UNC-83 UNC-84 LINC complex leads to displaced body wall muscle nuclei in C. elegans
Figure 1. A non-functioning UNC-83/UNC-84 LINC complex leads to displaced body wall muscle nuclei in C. elegans.

(A) Fluorescent images showing GFP-positive body wall muscle nuclei in WT (BAT661), unc-83(bar18) mutant (BAT1298, without the extrachromosomal rescue construct barEx453) and unc-83(bar18) mutant rescue animals (BAT1298, with the extrachromosomal rescue construct barEx453, driving WT unc-83 from the muscle-specific promoter myo-3p). In the rescued animal, the pharynx expresses myo-2p::mCherry which is marking transgenic animals carrying barEx453. Dashed boxes highlight the area around the posterior pharyngeal bulb, where nuclei accumulate in the mutant. Anterior is to the left.

(B) Phenotype penetrance of body wall muscle nuclei mis-positioning penetrance in WT (BAT661), unc-83(bar18) mutant (BAT197), unc-83(ku18) mutant (BAT1980), unc-84(e1174) mutant (BAT968), unc-83(bar18) muscle-specific rescued (BAT1298, BAT1906, BAT1907) and unc-83(bar18) ubiquitously rescued (BAT1300, BAT1908, BAT1909) animals. n ≥300 for each sample; 3 biological repeats; Error bars represent SEM.

(C) Amount of GFP-positive nuclei (excluding sex muscles) in WT (BAT1488) and unc-83(bar18) mutant (BAT1099) animals at the larval stages L1, L3 and L4. For L1 and L3 stages, these are exclusively body wall muscle cells. For the L4 stage, nuclei of other myo-3p::gfp-positive muscles, like somatic sheath or enteric muscles, contribute to the overall number of counted cells. n ≥10 for each condition; n.s. = not significant according to a student’s t-test (p >0.05) between WT and mutant; Error bars represent SEM.

(D) Differential interference contrast (DIC) microscopic image of a L3 worm. The defined areas head (anterior part of the worm until after anterior pharyngeal bulb), neck (end of head region until after first pair of intestinal nuclei) and posterior body (end of neck region until posterior end of the worm) are highlighted. The differences in body wall muscle nuclei numbers between WT (BAT1488) and unc-83(bar18) mutant (BAT1099) worms, are shown (based on the values shown in (E)).

(E) Distribution of body wall muscle nuclei in WT (BAT1488) and unc-83(bar18) mutant (BAT1099) L3 animals. There is no significant difference for the total number of body wall muscle nuclei. For the head and posterior body region, mutant animals display reduced amounts of body wall muscle nuclei, while the number of nuclei in the neck region is increased. Statistical significance based on student’s t-test. n =12 for each condition; *p ≤0.02, ***p ≤0.0001; Error bars represent SEM.


The nuclear lamina is connected to the cytoskeleton via different ‘Linker of Nucleoskeleton and Cytoskeleton’ (LINC) complexes with a variety of functions. LINC complexes are widely conserved over various phyla, which include organisms such as plants, slime molds, yeast, roundworms, fruit flies and mammals. LINC complexes cross the nuclear membrane and are composed of SUN and KASH domain-containing proteins, which interact in the perinuclear space between the inner and outer nuclear membrane.

KASH proteins are located at the outer nuclear membrane and may interact with actin filaments, microtubules (via dynein and kinesin), intermediate filaments (via spectrin), centrosomes and other cytoplasmic organelles. SUN proteins are located at the inner nuclear membrane and are associated with both chromatin and nuclear lamins. Functions include nuclear movement and anchoring, moving meiotic chromosomes and telomeres and sensing mechanic stimuli[1][2][3].

The KASH protein UNC-83 and the SUN protein UNC-84 form a LINC complex in C. elegans, which is required for migration of nuclei in P cells, intestinal cells and hyp7 hypodermal precursors, by recruiting dynein and kinesin-1 to the nuclear surface[4][5][6][7][8][9]. Furthermore, UNC-84 has been implicated in maintaining the nuclear architecture of force-bearing cells, like body wall muscles[10].


Our aim was to describe and quantify the novel observation of mis-positioned body wall muscle nuclei upon loss of the UNC-83/UNC-84 LINC complex and to address the question of whether the effect was cell autonomous or not.

Results & Discussion

We performed a forward genetics EMS screen to isolate mutants with an aberrant number of muscle nuclei in the nematode C. elegans. After mutagenizing worms carrying the muscle specific myosin reporter construct ccIs4251 [myo-3p::gfp::NLS] I, the F1 generation was analyzed for an atypical number of GFP-positive nuclei. A semi-automated high-throughput screen using a system that allows fluorescence-assisted sorting of large particles (Biosorter, Union Biometrica), yielded the isolation of a single mutant. This mutant was termed bar18 and showed an accumulation of muscle nuclei around the posterior pharyngeal bulb (Fig. 1A).

To identify the relevant mutation, we used whole-genome sequencing in conjunction with a SNP Mapping Strategy[11] and a published CloudMap pipeline[12]. We identified a premature STOP in the KASH-domain containing gene unc-83. To test whether this mutation was causing the observed phenotype, we performed rescue experiments, driving WT unc-83 from an extrachromosomal array either under the control of the muscle specific promoter myo-3p or the ubiquitous promoter eft-3p (Fig. 1A, 1B).

Since driving WT unc-83 from either promoter rescues this phenotype, we could not only confirm that the phenotype-causing mutation bar18 indeed belongs to unc-83, but could also show that the effect is cell autonomous. Furthermore, we could phenocopy the dispositioning effect using the unc-83(ku18) premature STOP allele, which supports our rescue experiments (Fig. 1B). In addition, we tested the unc-84(e1174) deletion allele, which also shows this phenotype, supporting the assumption that a non-functioning UNC-83/UNC-84 LINC complex is responsible for the mis-positioning of muscle nuclei. Alleles are summarized in supplementary figure 1.

Next, we quantified the amount of GFP-positive muscle nuclei in WT and unc-83(bar18) worms at different larval stages using the reporter construct rrrSi261 [myo-3p:: gfp::H2B] I (in contrast to the reporter described above, a single copy of rrrSi261 is integrated into the genome thereby making it very dim, but more stable without any tendency for mosaicism).

Surprisingly, the overall number of GFP-positive cells remained unchanged for all larval stages (Fig. 1C). At L1 stage, we counted 78.5 vs. 78.6 nuclei, at L3 stage we counted 94.7 vs. 94.6 nuclei and at L4 we counted 106.6 vs. 105.3 nuclei in unc-83(bar18) or WT, respectively. Our results are comparable with previously reported numbers of muscle cells[13]: C. elegans has an invariant number of somatic cells, including 95 striated body wall muscles from the L2 stage onwards (81 in L1) and several other non-striated muscles, some of which are born at later larval stages.

In order to quantify the positioning defect of body wall muscle nuclei in unc-83(bar18) animals, we sub-divided animals into 3 different regions (head, neck, posterior body) and quantified the amount of body wall muscle nuclei in WT and unc-83(bar18) animals (Fig. 1D, 1E). The neck region was defined as the region between the anterior pharyngeal bulb and the first pair of intestinal nuclei. The anterior region was defined as head and posterior was defined as posterior body (Fig. 1D).

Overall, unc-83(bar18) animals displayed 36.1% less nuclei in the head region (7.8 vs. 12.3), 28.7% more nuclei at the neck region (28.3 vs. 22.0) and 3.1% less nuclei in the remainder region 58.5 vs. 60.3). Our findings suggest that the nuclei accumulating in the neck region of unc-83 mutants originate primarily from the head region (Fig. 1D, 1E).


We describe a so far uncharacterized phenotype of mis-positioned body wall muscle nuclei upon lack of a functioning UNC-83/UNC-84 LINC complex in C. elegans. Unc-83/84 mutant animals display an accumulation of body wall muscle nuclei around the posterior pharyngeal bulb. Our data suggests that this cell autonomous effect is primarily due to nuclei that are displaced from the head of the worm towards the neck region. Our findings broaden our current understanding of the ubiquitously expressed LINC complex, which was so far described to ensure proper nuclei positioning in P cells, the intestine and hyp7 hypodermal precursors, but not for muscle tissue.


The reporters we used to show the nuclei mis-positioning effect in unc-83(bar18) animals are nuclear, so they cannot distinguish between mis-positioning of whole cells or of nuclei only. Since the somatic muscle tissue in C. elegans is not syncytial (unlike the epidermis), it’s likely that whole cells are mis-positioned, but the current study does not address this question.

Furthermore, we did not analyze whether these mis-positioned nuclei have any physiological impact on the worm. For example, if they are linked to previously observed pleiotropic phenotypes[14] such as Unc (uncoordinated, impaired movement) or Egl (egg laying defect).

Finally, it remains to be determined when in development the defect occurs. Since we can rescue the mis-localization of body wall muscle nuclei by driving the WT gene from a myo-3 promoter, it is likely that the defects are manifested after the myo-3 promoter gets activated during embryonic development.

Alternative Explanations

We explain our observation of accumulating body wall muscle nuclei in unc-83 mutants around the posterior pharyngeal bulb as being a result of nuclear mis-positioning. An alternative explanation would be that some cells in the head region are missing, while some muscle cells in the neck region undergo additional cell divisions, thereby leading to an accumulation of nuclei (and cells) in the neck area. While we think this scenario is rather unlikely and suggest a similar nuclear migration mechanism as it was shown for UNC-83/84 LINC in other lineages, we cannot fully exclude that additional cell divisions might occur.


Our next steps will be adressing the questions that we have outlined in the ‘Limitations’ part.


Nematode culture and crossings

C. elegans strains were maintained on nematode growth medium (NGM) seeded with OP50 bacteria at 15°C as described previously[15]. This temperature was chosen in preference to 20°C, since for unc-83 and unc-84 mutants, elevated temperatures were reported to increase a p-cell nuclei migration defect[8] as well as an egg laying defect[14], which we thought best to avoid, since our observed phenotype is temperature independent. To create BAT1099 and BAT1488, PMW200 was crossed with BAT197 and BAT661 (Strains list see Suppl. Table 1).

EMS mutagenesis

Synchronized young L4 larvae (BAT60) were harvested and resuspended in 1 ml M9 buffer (approx. from three 6 cm dishes). 7 µl of EMS (Ethyl methanesulfonate, Sigma #M0880) was dissolved in 1 ml of M9 and combined with the 1 ml of worms. EMS-treated worms were incubated on a rotary shaker for 4 h at RT, washed 3 times with M9 and finally plated on NGM-plates. After a 30 min recovery, 20 healthy late L4s were transferred to fresh plates (5 P0 animals per plate). F2 generations were screened for an aberrant number of GFP-positive nuclei using a Biosorter (Union Biometrica) and candidates subsequently verified with a fluorescent dissecting microscope. The resulting positive candidate BAT173 was crossed with N2 worms to generate BAT197.

Identification of the mutated bar18 locus

To identify the mutated bar18 locus, we applied a Hawaiian SNP crossing strategy and combined it with whole-genome sequencing (WGS) as described by Doitsidou et al.[11]. Using this strategy we introduced Hawaiian SNPs through crossing and homologous recombination into the genome of the mutant and then narrowed down the location of the mutation following WGS. After each cross, only animals that still show the mutant phenotype are analyzed and, as a result, fewer or no Hawaiian SNPs will be found around the bar18 locus, making it possible to identify it.

Briefly, we crossed bar18 mutant (BAT173) and WT (BAT60) animals with Hawaiian CB4856 males and singled F1 cross progeny. After they self-fertilized and propagated, we singled 41 independent F2s of bar18 worms that showed the nuclei displacement phenotype and 36 independent F2s of WT worms to fresh plates and allowed them to self-fertilize. After their progeny populated 6 cm NGM plates, worms were washed off with M9 and their DNA was isolated using the Gentra Puregene Tissue Kit (Qiagen, #158689) according to the manufacturer’s instructions.

Pooled DNA samples of bar18 and WT worms were used to prepare libraries with the Paired-End Sample Prep Kit (Illumina #PE-102-1002) according to the manufacturer’s protocol. Libraries were subjected to WGS on an Illumina HiSeq 2500 System, using single 100 nucleotide reads. Bioinformatic analysis was done using the Galaxy-based CloudMap tool according Minewich et al.[12].

Generation of transgenic rescue lines

To generate BAT1298 (myo-3p rescue) and BAT1300 (eft-3p rescue), the plasmids dBT599 (myo-3p::unc-83_cDNA::SL2::NLS::tagRFP::tbb-2_3’UTR) and dBT573 (eft-3p::unc-83_cDNA::SL2::NLS::tagRFP::tbb-2_3’UTR) were linearized (ScaI) and injected into hermaphrodites as complex arrays at a final concentration of 25 ng/µl or 10 ng/µl respectively, together with 4 ng/µl of a linearized (ScaI) hygromycin resistance plasmid (IR98, a gift from Sebastian Greiss) and 2.5 ng/µl of linearized (FspI) myo-2p::mCherry co-injection marker (pCFJ90, addgene plasmid #19327). Digested (PvuII) bacterial genomic DNA was added to a final DNA concentration of 150 ng/µl.

To generate BAT1906, BAT1907 (both myo-3p recue), BAT1908, BAT1909 (both eft-3p rescue), the plasmids dBT742 (myo-3p::3xFLAG::unc-83_cDNA) and dBT573 (eft-3p::3xFLAG::unc-83_cDNA) were linearized (ScaI) and injected into hermaphrodites as complex arrays at a final concentration of 5 ng/µl, together with 4 ng/µl of a linearized (ScaI) hygromycin resistance plasmid (IR98, a gift from Sebastian Greiss) and 5 ng/µl of linearized (ApaI) ttx-3p::mCherry co-injection marker (dBT158). Digested (PvuII) bacterial genomic DNA was added to a final DNA concentration of 150 ng/µl.
Sequences, plasmids and worm strains are available upon request.

Microscopic analysis

To quantify the nuclei dispositioning phenotype penetrance (Fig. 1B), living worms were assessed under a fluorescent dissecting microscope. To count total numbers of GFP-positive nuclei, worms were immobilized with polystyrene microspheres as described by Fang-Yen et al.[16]. We took image stacks of whole worms using a Leica DM6B-Z microscope. Stacks were opened using Fiji (http://fiji.sc) and nuclei quantified with the help of the multi-selection tool to mark counted nuclei. Sex muscles (vulva and uterine muscles) were excluded when counting.

List of worm strains used

(See Suppl. Table 1.)

Funding Statement

This work was sponsored by the ERC-StG-2014-637530 and ERC CIG PCIG12-GA-2012-333922 and is supported by the Max Delbrueck Center for Molecular Medicine in the Helmholtz Association.


We thank Sergej Herzog, Anne Krause and Kitty van Scharenburg for technical assistance and Anna Reid for her help to refine the manuscript. We thank members of the Tursun group for comments and critical discussions, Sebastian Greiss for the plasmid IR98, Peter Meister for the PMW200 worm strain and CGC (supported by the NIH) for other worm strains.

Conflict Of Interest

The authors declare no conflicts of interest.

Ethics Statement

Not Applicable.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.