(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Nuclear lipid droplets and nuclear damage in Caenorhabditis elegans

['Jose Verdezoto Mosquera', 'Fred Hutchinson Cancer Research Center', 'Seattle', 'Washington', 'United States Of America', 'Meghan C. Bacher', 'James R. Priess', 'Molecular', 'Cellular Biology Program', 'University Of Washington']

Date: 2021-08

Fat stored in the form of lipid droplets has long been considered a defining characteristic of cytoplasm. However, recent studies have shown that nuclear lipid droplets occur in multiple cells and tissues, including in human patients with fatty liver disease. The function(s) of stored fat in the nucleus has not been determined, and it is possible that nuclear fat is beneficial in some situations. Conversely, nuclear lipid droplets might instead be deleterious by disrupting nuclear organization or triggering aggregation of hydrophobic proteins. We show here that nuclear lipid droplets occur normally in C. elegans intestinal cells and germ cells, but appear to be associated with damage only in the intestine. Lipid droplets in intestinal nuclei can be associated with novel bundles of microfilaments (nuclear actin) and membrane tubules that might have roles in damage repair. To increase the normal, low frequency of nuclear lipid droplets in wild-type animals, we used a forward genetic screen to isolate mutants with abnormally large or abundant nuclear lipid droplets. Genetic analysis and cloning of three such mutants showed that the genes encode the lipid regulator SEIP-1/seipin, the inner nuclear membrane protein NEMP-1/Nemp1/TMEM194A, and a component of COPI vesicles called COPA-1/α-COP. We present several lines of evidence that the nuclear lipid droplet phenotype of copa-1 mutants results from a defect in retrieving mislocalized membrane proteins that normally reside in the endoplasmic reticulum. The seip-1 mutant causes most germ cells to have nuclear lipid droplets, the largest of which occupy more than a third of the nuclear volume. Nevertheless, the nuclear lipid droplets do not trigger apoptosis, and the germ cells differentiate into gametes that produce viable, healthy progeny. Thus, our results suggest that nuclear lipid droplets are detrimental to intestinal nuclei, but have no obvious deleterious effect on germ nuclei.

Several human disorders such as obesity are associated with abnormal fat storage. Cells normally store fat in cytoplasmic organelles called lipid droplets. However, recent studies have shown that fat can also form inside of the cell nucleus, and the effects of nuclear fat are not known. Here we use the cell biology and genetics of the model organism C. elegans to study the causes and consequences of nuclear fat. We show that intestinal cells can contain nuclear fat, particularly during high-low-high changes in cytoplasmic fat that involve de novo fat synthesis. Nuclear fat is associated with multiple changes in intestinal nuclei that appear to represent damage and repair. Germ nuclei that normally differentiate into oocytes can also contain nuclear fat. In germ cells, however, even high levels of nuclear fat appear to cause little or no damage. Our results suggest that intestinal nuclei and germ cell nuclei might have different responses to nuclear fat in part because they differ in chromosomal organization at the nuclear envelope.

We show that nLDs occur in normal intestinal nuclei, and are associated with complex changes in nuclei during the self-fertile period. These nuclear changes include the development of bundled microfilaments, the formation of a type I nucleoplasmic reticulum, and the generation of lamin-coated vesicles. We show that a few intestinal nuclei appear to rupture at sites adjacent to large nLDs, and that some nLDs appear to initiate a series of events that culminate in the formation of large nuclear cysts and the removal of aberrant nucleoplasm. These nuclear changes occur predominantly in a subset of adult intestinal cells that have the greatest fluctuation in stored fat during the self-fertile reproductive period. We show that germ cell nuclei also form nLDs, but the nuclei appear relatively unchanged. Because of the simplicity of nLD formation in germ cells, we used the gonad in a forward genetic screen for mutants with abnormal numbers or sizes of nLDs. Molecular analysis of four mutations showed that they affect the inner nuclear membrane protein NEMP-1/NEMP1, the lipid droplet regulator SEIP-1/seipin, and the COPI proteins COPA-1/α-COP and COPB-2/β’-COP. Surprisingly, these mutants suggest that a large volume of a meiotic germ nucleus can be filled with fat without triggering apoptosis, and that embryos derived from such germ cells are viable and healthy. Our results suggest that the nLD-associated defects in intestinal nuclei, and the lack of similar defects in germ cells, result from structural differences in the two types of nuclei.

During the self-fertile period, unmated hermaphrodites use their stored sperm to produce limited numbers of self-progeny, and egg production ceases once those sperm are depleted. However, hermaphrodites maintain the potential for producing far larger numbers of cross-progeny if they encounter males and mate after the self-fertile period. Thus, any nuclear defects in the self-fertile period would likely result from specific trauma, and be repaired, rather than from a generalized deterioration associated with post-reproductive senescence. The hermaphrodite gonad provides a sensitive system for addressing whether physiological or cellular changes are deleterious, because germ cells can be triggered to undergo apoptosis by a large variety of stresses that include DNA damage, infection, heat, and starvation [ 35 ]. Indeed, more than 50% of germ cells undergo apoptosis in wild-type animals grown under standard culture conditions; the mechanisms that trigger this natural or "physiological" apoptosis are largely unknown [ 29 , 36 – 38 ]. Importantly, mutations that block germ cell apoptosis are homozygous viable, allowing possible defects to be analyzed in the surviving progeny. Although intestinal cells lack an apoptosis pathway, they undergo necrosis in response to various traumas such as bacterial infection or hypo-osmotic shock [ 39 , 40 ].

The goals of the present study were to address whether nLDs are (1) present in selected C. elegans tissues, and (2) drive nuclear damage. In particular, we hoped to determine whether nLDs might be a factor in the extraordinary number of apoptotic deaths that occur in normal germ cell development [ 29 ]. Most studies on nLDs in cultured cells induce nLD formation with oleic acid-supplemented media. Here, we analyzed two fatty tissues, the intestine and the gonad, in animals fed the standard laboratory diet of E. coli strain OP50. The intestine is the major fat-storage tissue in C. elegans; it has been studied extensively as a genetic model of fat storage and obesity, and compared with both liver and adipose tissue in higher animals [ 30 – 32 ]. We focused on the self-fertile reproductive period when adult hermaphrodites naturally undergo large changes in fat; these changes involve yolk production by the intestine, and egg production by the gonad. Larval hermaphrodites produce only low-fat sperm, but adult hermaphrodites switch to producing high-fat oocytes. The spermatogenesis/oogenesis switch is associated with a massive reallocation or transfer of metabolic resources, such as fat precursors and yolk lipoproteins, from the intestine to the gonad. Those resources are in turn lost from the gonad as eggs are laid. Remarkably, the peak amount of eggs laid in one day is equivalent to the entire body weight of the adult, and just one yolk protein, YP170, accounts for about 25% of protein synthesis in the intestine [ 33 , 34 ].

nLDs that form at the envelope presumably enter the nucleoplasm by penetrating the nuclear lamina and associated peripheral heterochromatin [ 24 , 25 ]. Thus, nLDs have the potential to impact chromatin organization or the integrity of the lamina. Lamins are major components of the lamina, and interact directly or indirectly with proteins that have diverse roles in nuclear biology, including nuclear architecture and chromatin organization. C. elegans has a single lamin, LMN-1, which is most similar to B-type lamins in other systems [ 26 , 27 ]. Mutations in lamin and other lamina components cause severe developmental defects and shortened lifespan in C. elegans, and are responsible for numerous human diseases including muscle dystrophies, lipodystrophies and premature aging (reviewed in [ 28 ]).

LDs are often found in the cytoplasm adjacent to the outer nuclear membrane, which is continuous with ER membranes [ 21 ]. However, several recent studies have shown that LDs also occur inside nuclei [ 22 ]. The function(s) of nuclear lipid droplets (nLDs) are not known, but possibilities include providing lipid for the growth of nuclear membranes, storage sites for normal proteins or for unfolded, hydrophobic proteins, and as sites for the detoxification of hydrophobic substances [ 23 ]. Few or no nLDs have been observed in some types of cells that contain abundant cLDs, such as adipocytes, or stimulated fibroblasts, but nLDs have been found in multiple types of liver-derived cell lines [ 24 ] and in human patients with fatty-liver disease, or hepatic steatosis [ 22 ]. Obesity-related hepatic steatosis is estimated to affect 20–30% of the population of North America, including very high percentages of patients with morbid obesity or type 2 diabetes (reviewed in [ 20 ]). Although the normal mammalian liver has few cytoplasmic LDs (hereafter cLDs) in the fed state, even short periods of fasting cause an enormous accumulation of cLDs, as lipids are mobilized from adipose tissue and stored in liver cells to reserve energy for vital functions.

LDs consist of neutral lipids, largely triacylglycerides and sterol esters, that are surrounded by a surface monolayer of polar phospholipids. LDs can grow by fusing with other LDs, but de novo formation occurs in the endoplasmic reticulum (ER) [ 8 , 9 ]. Lipid appears to accumulate among the acyl chains on the lipid bilayer of the ER membrane. The lipid coalesces and expands into a stable "lens" that separates the two leaflets of the bilayer, eventually budding into the cytoplasm as a small LD. Homooligomers of the protein seipin appear to form toroids around ER-LD contact sites [ 10 – 12 ], and function in stabilizing the lipid lens and in trafficking protein/lipid into the developing LD [ 9 , 13 ]. Loss-of-function mutations in human seipin cause BSCL type 2, one of the most severe lipodystrophies, while gain-of-function mutations in seipin cause pathologies of the nervous system [ 14 , 15 ]. Among the many known regulators of lipid droplets, COPI (coat protein complex I) components appear to have a role in LD morphology, protein composition, and lipolysis [ 16 – 18 ]. For example, COPI-dependent membrane bridges appear to link LDs with the ER, allowing ER-localized regulatory proteins to load onto the LD surface [ 19 ]. Some ER-localized enzymes involved in triglyceride synthesis appear to remain anchored in the phospholipid monolayer of the LD surface after budding, and drive the continued growth of the LD [ 20 ].

Lipid droplets (LDs) are the major form of fat storage, and provide a reservoir of energy and building blocks for membrane growth and repair. LDs can have additional, important functions in cell signaling, development, and in stress responses [ 1 , 2 ]. For example, lipid droplets can serve as storage sites for histones in Drosophila embryos, and can contain polyubiquitinated proteins in yeast cells that experience acute lipid stress [ 3 , 4 ]. Excess lipid accumulation is associated with a wide range of human pathologies from atherosclerosis to obesity [ 5 ]. At the cellular level, dysfunction in lipid storage or metabolism can lead to free fatty acid-induced lipotoxicity, with cascading trauma to cell membranes and other components [ 6 ]. Many types of cells contain LDs, although the numbers and sizes of LDs can vary enormously depending on physiological state or culture conditions. For example, the volume of stored lipid in cultured cells can rapidly change severalfold upon fatty acid supplementation or withdrawal [ 7 ].

Results

Background C. elegans undergoes four larval stages, called L1-L4, before becoming an adult. The nomenclature used to indicate adult ages varies in the literature; here, a Day 1 (D1) adult was picked as a mid-L4 larva and allowed to develop for 24 hours (Fig 1A). This study focuses on unmated hermaphrodites during the self-fertile reproductive period, between L4 and D4 (Fig 1A). The mean lifespan is D16 with the standard diet and culture conditions used here, so D1-D4 adults represent the first quarter of adult life [41]. PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 1. Changes in intestinal fat during the self-fertile reproductive period. (A) Timeline of C. elegans development at 20°C, showing key events (adapted from [111]). (B) Anterior half of a D1 intestine imaged by Differential Interference Contrast (DIC) microscopy after staining for fat (red, oil red O); two intestinal nuclei are outlined in white. Note that the intestine has a much higher level of fat than the pharynx. The intestine in a newly hatched larva consists of groups of cells called int1-int9, and additional cells are added during post-embryonic development. This report defines the term subanterior to refer collectively to cells in the anterior half of the intestine, but excluding the anteriormost, int1 group. (C) Comparison of fat levels in hermaphrodite tissues. Note the large difference in fat between the L4 and adult gonads, which produce sperm and oocytes, respectively. (D) Images of D2 intestines stained for fat with oil red O (red, top panel) or with BODIPY (green, bottom panel). Both stains show a relative reduction of fat in the subanterior region, with the greatest variation in the int2 cell group. (E) High magnification of cells in a D2 intestine, comparing fat (green, BODIPY) in cells in the int1 group, cells in the subanterior region, and cells in the posterior region. For contrast, the DAPI-stained nuclei are shown in red. Each image is a projection of a 5 μm optical z-stack. (F) Comparison of intestinal fat in animals at the indicated stages. Approximately equal numbers of anterior and posterior dissections were performed as quantified in Fig 1G, but posterior dissections are only shown for the D2 timepoint. Note that fat is depleted, then replenished, in the subanterior region. (G) The chart shows the level of fat in the subanterior region compared with the flanking regions (the int1 group and the posterior cells); data is shown for both hermaphrodites and males. The total numbers of hermaphrodite/male intestines examined were L4 (48/17), D1 (37/21), D2 (52/14), D3 (41/22) and D4 (47/16). (H) Comparison of fat in the subanterior region of staged intestines as indicated. (I) Comparison of fat in D4 mated animals (left) with synchronous, unmated D4 controls (right). For this experiment, males were mated with D2 hermaphrodites for 24 hours and then removed. Mated D4 hermaphrodites were identified as containing fertilized eggs. (J) Candidate nLDs (arrows) in the subanterior region of D2 intestines. Note that the candidate nLDs are far larger than cLDs in the surrounding cytoplasm. Nuclei are indicated by dashed lines. Scale bar sizes are shown in microns for all panels. https://doi.org/10.1371/journal.pgen.1009602.g001 During feeding, ingested bacteria are crushed by a muscular pharynx, then passed through a small epithelial valve and into the intestine for digestion (Fig 1B). The 30–34 nuclei in the adult intestine undergo endoreduplication to reach a ploidy of 32C, and are much larger than most other C. elegans nuclei [42]. Cells in the intestine form natural groupings; the anteriormost four cells are called the int1 group, and more posterior groups consist of two cells each.

The intestine undergoes region-specific changes in fat during the self-fertile period We examined fat in aldehyde-fixed intestines and gonads stained with the lipid dyes oil red O or BODIPY. L4 gonads, which produce low-fat sperm, had very low levels of fat, as expected (Fig 1C). D1 adult hermaphrodite gonads, which produce high-fat oocytes, were larger and contained a much higher level of fat (Fig 1C). All L4 intestines had a high, uniform level of fat (Fig 1C), as did about half of D1 intestines (Fig 1B). Unexpectedly, several D1 intestines and most D2 intestines had a non-uniform distribution of fat: Fat levels were high in the anteriormost cells (int1 group), and in all cells in the posterior half of the intestine (Fig 1D). However, fat levels were much lower in the region between the anterior and posterior cell groups, here termed the subanterior region (Fig 1D). Fat depletion in the subanterior region was associated with a large, relative decrease in the sizes and numbers of cLDs (Fig 1E). We next examined fat in synchronous populations of L4-D4 animals, covering the self-fertile period. Animals were dissected near the head to release the anterior half of the intestine (int1 cells plus the subanterior region), or near the tail to release the posterior half of the intestine (Fig 1F, quantified in 1G). We found that fat depleted in the subanterior region of D2 intestines appeared to be restored to high levels between D3 and D4 (Fig 1F–1H), as the self-sperm of unmated hermaphrodites are depleted and egg production ends. We next examined mated hermaphrodites that continue egg production beyond D4. The mated D4 hermaphrodites did not restore fat in the subanterior region, and showed additional depletion of fat in the flanking regions (Fig 1I). We next used transmission electron microscopy (TEM) to examine intestinal cell cytoplasm in L4, D2, and D3 hermaphrodites (S1 Fig). As expected, only the subanterior cells showed a large loss of cLDs between L4 and D2, followed by an increase at D3. However, all intestinal cells showed major changes in cytoplasmic components after the L4 stage, including a loss of glycogen and a decrease in the sizes and numbers of yolk granules. Thus, all intestinal cells appear to mobilize resources to support egg production, but only the subanterior cells undergo large, high-low-high changes in fat.

nLDs can be associated with nuclear microfilament bundles Some of the large nLDs with heterochromatin coats contacted, or were partially wrapped by, linear structures that stained positively for lamin (Fig 3A); for convenience we refer to these nuclear structures as lamin lines. L4 nuclei typically had few or no lamin lines, but the lines were common in older nuclei (Table 2); some individual D3 nuclei contained at least 9 lines (see below). All of the lamin lines were similar in thickness, but varied markedly in length; the longest lines were about 10 μm, or about 2/3 of the nuclear diameter. In most cases, one end of a lamin line could be traced to the nuclear envelope (Fig 3A–3D). Remarkably, many of the lamin lines were surrounded by heterochromatin, particularly in older nuclei (Fig 3A and 3B). For example, heterochromatin was associated with nearly 20% of lamin lines in D2 nuclei, and 51% of the lines in D3 nuclei (n = 76, 80 lines, respectively). In addition to contacting nLDs, the lamin lines often contacted lamin-coated vesicles, or lamin sacs; some of the lamin sacs contained lipid, but other did not (Fig 3C and see below). Intranuclear lines of lamin have been described in other systems and found to be invaginations of the nuclear envelope [43]. However, the lamin lines in intestinal nuclei were not surrounded by nuclear pores, suggesting they are not simply nuclear invaginations (Fig 3D, n = 0/76 lines). PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 3. nLDs and nuclear microfilament bundles (nMFBs). (A) Examples of lamin lines (arrowheads; red, LMN-1/lamin) and nLDs (green, BODIPY) in two D1 nuclei. The lamin line in panel 1 extends between the envelope (arrows) and the nLD (top inset). The bottom inset in panel 1 shows that both the nLD and the lamin line are coated with heterochromatin. The nucleus in panel 2 has three lamin lines, one of which partially encircles the nLD; the top inset shows the continuation of the latter lamin line on a different focal plane. (B) D1 nucleus with two lamin lines (arrowheads), both coated with heterochromatin. (C) D2 nucleus with a lamin line (arrowhead) in contact with a lamin-coated vesicle. (D) Lamin line (arrowhead) in a D2 nucleus stained for nuclear pores (green, NPP-9/RanBP2). Note that the line is not surrounded by nuclear pores, suggesting that the line does not represent an invagination of the nuclear envelope. (E) Low magnification TEM of a D3 nucleus showing a linear structure extending from the nuclear membrane into the nucleoplasm. The structure is coated with clumps of electron-dense material (black arrowheads) consistent with the appearance of heterochromatin. The high magnification inset shows that the linear structure is a bundle of parallel microfilaments (nMFB; the white arrowhead indicates a single microfilament). (F) Examples of nMFBs in D1 (panel 1), D2 (panels 2,3) and D3 (panel 4) nuclei. Note that the nMFBs in panels 2 and 3 appear to connect directly to the INM (arrow). The nMFB can be coated with variable clumps of electron-dense material (black arrowheads), or be associated with membranes (black arrows in panel 4). (G) TEM of D2 nucleus showing an nMFB (arrowhead) in contact with a nLD. (H) TEM of D3 nucleus showing an nMFB (arrowhead) in contact with two vesicles and a tubule (arrow and inset). One of the vesicles is a kernel vesicle as described in the text. (I) Examples from D2 nuclei of tubules (arrows) extending parallel to nMFBs (arrowheads). (J) Examples of nMFB-associated tubules (arrows) that appear to be protrusions from membrane vesicles. Note that panel 1 also shows two nLDs that appear to be surrounded by additional membranes, as described below. Scale bars in microns as indicated. https://doi.org/10.1371/journal.pgen.1009602.g003 Intestinal nuclei examined by TEM often contained linear structures that closely resembled the lamin lines in size and age-dependence (Fig 3E). At high magnification, the linear structures consisted of bundled, parallel microfilaments that were the thickness expected for F-actin (about 6 nm; inset for Fig 3E). The bundles were not surrounded by a double membrane, indicating that they are not cytoplasmic microfilaments within a nuclear invagination (Fig 3F). Thus, we refer to the bundles as nMFBs (nuclear microfilament bundles). Similar to many lamin lines, the nMFBs often were associated with irregular clumps of electron-dense material resembling heterochromatin (black arrowheads in Fig 3E and 3F). The nMFBs appeared to originate/terminate at the INM (panels 2,3 in Fig 3F), and could contact nLDs (Fig 3G) or other nuclear vesicles as described below (Fig 3H).

A third class of nLDs is surrounded by lamin About 30% of the nLDs with heterochromatin coats in L4-D2 nuclei were also enclosed by a sac of lamin (Fig 6A and 6B). The lamin sacs could be filled entirely with lipid (Fig 6A), but often contained a much smaller lipid core, or rarely multiple lipid bodies (Fig 6B and 6C). None of the lipid bodies within lamin sacs had heterochromatin coats, although the sac itself was usually surrounded by heterochromatin (Fig 6C). By contrast with other nLDs in the nucleoplasm, the lamin sacs were nearly always adjacent to the nuclear envelope (see Legend to Fig 6D). The numbers of lamin sacs increased progressively between the D1 and D3 stages, and the sacs were most abundant in subanterior nuclei (Table 2 and arrowheads in Fig 6D). For example, about 80% of D3 subanterior nuclei had 5 or more lamin sacs, but only 10–30% of the flanking intestinal nuclei had similar numbers of sacs. The lamin sacs in older nuclei appeared similar in size and location to those in younger nuclei, but most did not appear to contain lipid (panel 3 in Fig 6B, 5/64 sacs with lipid). Thus, the sacs could either lose lipid over time, or sacs in older nuclei might form independent of lipid. Nearly all of the lamin sacs were at gaps in the peripheral heterochromatin (arrowheads in Fig 6E). The sacs in D1 nuclei typically had a distinct coat of heterochromatin, but most sacs in D3 and D4 nuclei had little or no coating (Fig 6E). PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 6. nLDs and lamin sacs. (A) Examples of nLDs with and without lamin coats (red, LMN-1/lamin) in L4 (panel 1) and D1 (panels 2,3) intestinal nuclei. (B) Examples of nLDs that are within, but that do not fill, lamin sacs (arrows). The lamin sacs are closely associated with the nuclear envelope; panels 2 and 3 are tangential planes through the surfaces of D2 and D3 nuclei, respectively. Panel 3 shows a 5 μm maximum intensity z projection to illustrate the number of lamin sacs; note that only one lamin sac appears to contain lipid (arrow). (C) High magnification views of nLDs (arrows) within lamin sacs. The small, internal lipid droplets (arrows) do not appear to be directly surrounded by heterochromatin, although the lamin sac itself has a variable coating of heterochromatin (arrowheads). (D) Intestinal nuclei stained for lamin (white, LMN-1) from different regions of the intestine and at different stages as labeled; the images are maximum intensity projections showing the entire nucleus. Lamin lines (arrows) in the projection are in the nuclear interior, but nearly all of the lamin sacs (arrowheads) are just inside the nuclear envelope. (E) Images of lamin sacs in D1 and D3 nuclei. Lamin sacs in D1 nuclei have distinct coatings of heterochromatin that appear to result from inpocketings of the peripheral heterochromatin. Lamin sacs in D3 and older nuclei typically have relatively little, if any, coating of heterochromatin. (F) D2 nucleus showing a candidate precursor of a lamin sac. The image appears to show an inpocketing of the lamina (double arrow) and peripheral heterochromatin (arrows). (G) TEM of a D2 nucleus showing a candidate precursor of a lamin sac/kernel vesicle. The INM (double arrow) appears to have separated from the ONM to form an inpocketing, and there are no nuclear pores (arrowhead) at the base of the inpocketing. The membrane inpocketing has an irregular inner lining, and clumps of material appear at the outer, nucleoplasmic surface of the inpocketing. (H) D1 intestinal nuclei showing gaps (arrowheads) in the distribution of nuclear pores (green, NPP-9/RanBP2) that are coincident with the positions of lamin sacs. Panel 1 is an optical cross-section of a nucleus, and panel 2 is a tangential plane through the surface of a nucleus. (I) TEM and diagram of a kernel vesicle, or presumptive lamin sac, at a gap between pores. The kernel vesicle appears to surround a small nLD, similar to immunostained images (Fig 6C). The membrane surface of a kernel vesicle can be coated with variable clumps of material (white arrowheads in left panel), and nearly always has an irregular, inner lining of material. (J) Intensity scan through a kernel vesicle, showing the increased electron density of the lipid-like core. (K) TEM comparing a kernel vesicle membrane with the adjacent nuclear membranes. The high magnification inset shows that each of these membranes has the typical sandwich appearance of a lipid bilayer. This morphology is consistent with the proposed origin of the vesicle membrane as an inpocketing of the INM (see Fig 6G). Scale bars as indicated in microns. https://doi.org/10.1371/journal.pgen.1009602.g006

TEM shows that lamin sacs are membrane vesicles with possible lipid cores Images of possible nascent sacs in D1 and D2 nuclei appeared to show inpocketings of both the lamina and the peripheral heterochromatin (Fig 6F). Similar inpocketings observed by TEM showed a separation of the INM from the ONM (double-headed arrow in Fig 6G). The separation between the INM and ONM would be incompatible with the structure of a nuclear pores (Fig 5C); accordingly, we found that nearly all immunostained lamin sacs were beside circular gaps in the otherwise uniform distribution of nuclear pores (Fig 6H): When the positions of lamin sacs and the gaps between pores were scored separately and then compared, 102/110 lamin sacs were at gaps, and 80/93 gaps were adjacent to lamin sacs. This nearly invariant correspondence, combined with the sizes and age distribution of the sacs, allowed us to identify fully-formed sacs as distinct membrane vesicles by TEM (Fig 6I). The vast majority of these vesicles contained a smaller inner structure we term the kernel (Figs 6I and 6J and S4). The kernel consists of a spherical, homogenous core of intermediate electron density similar to nLDs, and the core is surrounded by a variable, non-membranous coat (Fig 6J). For clarity, we refer to TEM images of the presumptive lamin sacs as kernel vesicles, and use the term lamin sacs for immunostained nuclei. Similar to lamin sacs, the kernel vesicles were spherical, about 0.5–1.0 μm in diameter, absent in L4 nuclei, juxtaposed to the nuclear envelope, and most abundant in subanterior nuclei. The kernel vesicles were surrounded by a lipid bilayer membrane (right panel, Fig 6K). The outer (nucleoplasmic) face of the vesicle membrane was often associated with variable clumps of electron-dense material, and the inner face typically had a loosely-associated lining of material (Fig 6G and 6I–6K). In addition to having possible lipid cores, a few kernel vesicles had an nLD, or occasionally multiple nLDs, at their perimeter (Fig 7A), as did immunostained lamin sacs (Fig 7B). Thus, nLDs might form infrequently from the INM-derived membrane surrounding a lamin sac/kernel vesicle. PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 7. nLDs and nucleoplasmic cysts. (A) TEM images of nLDs (black arrows) at the perimeter of kernel vesicles (arrowheads). Note that there are five nLDs around the kernel vesicle in panel 3. Note also nuclear tubules in panels 1 and 4 (white arrows). (B) D2 nucleus with two nLDs, one at the perimeter of a lamin sac (arrowhead) as in Fig 7A. (C) Examples of D3 nuclei with small nLDs (arrows) inside large nucleoplasmic cavities or cysts. (D) Immunostained (left column) and TEM images (right column) of D3 nuclei from the indicated regions of the intestine. Large nucleoplasmic cysts are visible in the subanterior nuclei, but not in the flanking nuclei (see also S5 and S6 Figs). By TEM, the cysts do not resemble secondary nucleoli (labeled in the subanterior nucleus), and do not stain for nucleolar markers (green, anti-fibrillarin). Double arrows in the subanterior nuclei indicate apparent pairings of cysts with lamin sacs/nuclear vesicles. (E) Examples of apparent pairings (double arrows) between cysts and lamin sacs in D3 nuclei. Note that lamin sacs, but not cysts, are adjacent to gaps in the peripheral heterochromatin. (F) Intestinal nuclei in live, D3 worms expressing one of two transgenic, fluorescent reporters for histones. The left panel shows GFP::HIS-2B (green; strain JM149), and the right panel shows HIS-24::mCherry (red; strain RW10062). Both reporters show histone-deficient regions in the nucleoplasm (arrows) consistent with the variable shapes and sizes of cysts or sac/cyst pairs. (G) TEM of D3 nucleus with three kernel vesicles, one of which is paired with a cyst. (H) TEM of D2 (panels 1,2) and D3 (panels 3–5) nuclei showing kernel vesicles paired with cysts. Panels 1 and 2 show kernel vesicle membranes continuous with cyst membranes. The cysts in panels 3 and 4 contain lipid droplets (compare with Fig 7C). The cyst in panel 5 has a cauliflower shape and appears to be fragmented. https://doi.org/10.1371/journal.pgen.1009602.g007

Intestinal nuclei appear to remove nLD-associated debris D3 and D4 nuclei, but not younger nuclei, often contained regions of aberrant nucleoplasm that were comparable in size and shape to cysts, and that were adjacent to degraded kernel vesicles (Fig 8A). However, the aberrant nucleoplasm was bordered by variable membrane fragments, rather than a continuous membrane (arrows, Fig 8A and inset). These features suggest that the aberrant nucleoplasmic regions represent lysed cysts. D3 and D4 nuclei contained several examples of what appeared to be membrane vesicles or protrusions that engulfed aberrant nucleoplasm, degraded kernel vesicles, and unidentified structures (black arrowheads in Fig 8B). Engulfed materials and vesicles were often in clusters that contained nMFBs and tubules (Fig 8C). Similarly, immunostained D3 and D4 nuclei contained clusters of nLDs, lamin sacs and other, much smaller bodies that were covered with lamin; these clusters typically occurred at concavities or flattened regions of the nucleus (arrowheads, Fig 8D). For example, 60.7% of D4 subanterior nuclei contained clusters with five or more lamin sacs (n = 56 nuclei). The percentages of D3 and D4 nuclei with at least one lamin line/nMFB were only moderately higher than for D2 nuclei (Table 2). However, individual D3 and D4 nuclei could have many more lamin lines than younger nuclei, and nearly all of the lines in older nuclei were coated with heterochromatin (Fig 8E). Finally, several D3 and D4 nuclei had variable, lamin-containing structures that projected, or were detached, from the nucleus, and a few of these structures appeared to contain DNA (panel 3 in Fig 8F). Similarly, TEM showed that D3 and D4 nuclei could have protrusions of the ONM that contained apparent remnants of membranous structures and other material (S4 Fig). PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 8. Degradation/engulfment of nuclear vesicles and nLDs. (A) TEM of D3 and D4 nuclei. Each nucleus contains a large region of atypical nucleoplasm that is partially surrounded by membrane fragments and that is adjacent to a membrane vesicle. The vesicles are identified as degraded kernel vesicles by their size, position, and by the gaps between nuclear pores (arrowheads) at the bases of the vesicles. (B) TEM images of D3 and D4 nuclei showing engulfment of various nuclear structures; the black arrowheads indicate candidate engulfing membranes. Panel 1 shows a membranous projection contacting a degraded kernel vesicle. Panels 2 and 3 show membrane-engulfed nucleoplasm; note the membrane tubule (arrow) in panel 2. Panels 4–8 show apparent engulfment of kernel vesicles or degraded kernel vesicles. Panels 9–10 show engulfed vesicles of undetermined identity. (C) TEM of D3 and D4 nuclei with vesicle clusters. Panel 1 shows a long nMFB associated with a large cluster of vesicles; the nMFB contacts what appears to be engulfed, aberrant nucleoplasm (white arrow and inset). Panels 2 and 3 show vesicle clusters that include nLDs, degraded kernel vesicles (white arrowheads), membrane tubules (white arrows), and several uncharacterized vesicles or membrane-engulfed nucleoplasm. (D) D3 nuclei with clusters of lamin sacs (red, LMN-1) at flattened or indented regions of the envelope (arrowheads). (E) D4 nucleus with at least 7 lamin lines (asterisks). The left panel is a 3 μm maximum intensity z-projection and the right panel is a single optical plane showing two lamin lines, both surrounded by heterochromatin. (F) D4 nucleus with protruding or detached, lamin-containing structures (arrows). Panel 3 is a higher magnification showing DAPI-staining material in one protrusion. Compare with TEM images in S4 Fig. https://doi.org/10.1371/journal.pgen.1009602.g008 The above results suggest that hermaphrodite intestinal nuclei accumulate damage by the end of the self-fertile period, and try to repair/remove this damage. Much of this damage appears to be associated with degraded lamin sacs/kernel vesicles, at least some of which once contained lipid. Finally, the nuclear damage occurs most frequently in the subanterior region of the intestine, which experiences the largest changes in fat and has the highest frequency of nLDs.

nLDs disappear in the major apoptotic region, but independent of apoptosis To characterize the distribution of germ cell nLDs, we stained gonads for lipid, the nuclear envelope, and for nucleoli. Optical z-stacks were acquired through the entire gonad, and scored in 50 μm divisions from distal to proximal (Fig 10G); automated counts of nucleoli were used as a proxy for the total number of germ cells, and nLDs were confirmed by analysis of the nuclear envelope. The positions of the first germ cells with nLDs varied between gonads, but the earliest examples were meiotic cells at early pachytene (division II in Fig 10H). Because nLDs form before germ cells begin to take up yolk lipoproteins (Fig 10A), we did not expect nLD formation to depend on yolk. Indeed, rme-2(b1008) mutants which fail to take up yolk [51] appeared to have normal numbers of nLDs (n = 24 gonads, compare Fig 10E with S7 Fig). The percentages of D1 and D2 germ cells with nLDs increased to a peak of about 13–15% during pachytene, then declined in later meiotic stages (Fig 10H). The post-peak decline in nLDs occurs in the major apoptotic region of the gonad, where the surviving germ cells increase in size and intercalate into a single row of oocytes (Fig 10G and 10A). Because our TEM analysis showed that apoptotic cells can have nLDs (Fig 11D), we addressed whether the post-peak decline resulted from apoptosis of nLD-containing germ cells. The caspases CED-3 and CSP-1 are both expressed in the gonad, although only CED-3 is essential for germ cell apoptosis [57]. We found that the distribution profile of nLDs in ced-3 mutants and in csp-1;ced-3 double mutants was similar to wild type, with a distinct peak followed by a decrease in nLDs (Fig 10I). Thus, these experiments show that nLDs disappear independent of apoptosis, presumably by lipolysis. However, they do not address whether germ cells with nLDs could be preferentially targeted for apoptosis in wild-type gonads.

nLD formation is associated with rapid oogenesis Although nLDs were present in all adult hermaphrodite gonads, which produce oocytes, nLDs were not detected in the vast majority of L4 gonads, which produce sperm (Table 3). Similarly, nLDs were not detected in the vast majority of D1 and D2 male gonads (Table 3). To determine whether hermaphrodite spermatogenesis, which precedes oogenesis, had a role in nLD formation, we examined fog-2(q71) mutant hermaphrodites that produce oocytes but never produce sperm [58]. fog-2 mutant D1 adults had about the same percentage of germ cells with nLDs as wild type (peak zone 11.3% versus 12.1%, respectively; Table 3). However, fog-2 D2 adults had a much lower percentage of nLDs than wild type (4.6% versus 15.1%; Table 3). Signals from sperm are known to stimulate oogenesis, and would be present in D2 wild-type hermaphrodites but absent in unmated, D2 fog-2 females [59,60]. Thus, we examined mated fog-2 mutants at D2, and found they had a similar percentage of nLDs as wild type (18.9% versus 15.1% respectively; Table 3). Thus, nLD formation is associated with oogenesis irrespective of prior spermatogenesis. However, sperm stimulates nLD formation in older adults, likely by stimulating oogenesis. We next examined the effect of low culture temperature on nLD formation, as low temperatures slow oogenesis and development in general. Indeed, shifting the culture temperature from 20°C to 15°C significantly decreased both the numbers and sizes of germ cell nLDs in D1 and D2 gonads (S7 Fig). PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Table 3. Germ cell nLDs. https://doi.org/10.1371/journal.pgen.1009602.t003

nLD mutants are defective in proteins localized to the INM or involved in lipid synthesis NEMP-1 is the sole C. elegans ortholog of vertebrate Nemp1 (also called TMEM194A), and is a conserved INM protein with five transmembrane domains (Fig 14A) [70–72]. The function of NEMP-1/Nemp1 is not known; it has been suggested to have roles in Xenopus neural development, tamoxifen resistance in human breast cancer cells, and to contribute to the stiffness of the nuclear envelope [70,72,73]. Nemp1 localizes with lamin, and appears to interact with lamin through a region spanning the five transmembrane domains (Fig 14A) [74]. The C-terminal half of Nemp1 extends into the nucleoplasm, and includes a domain that can bind RanGTP [70]. The semi-dominant nemp-1(zu501) mutation is predicted to encode a truncated membrane protein that lacks the RanGTP domain (Fig 14A). To confirm that the nemp-1(zu501) mutation was responsible for the nLD phenotype, we used Crispr-Cas9 to engineer identical (zu522) and similar (zu523) mutations, and found that both mutations resulted in apparently identical nLD phenotypes. SEIP-1 is the sole C. elegans ortholog of vertebrate seipin, an ER-resident membrane protein that functions in lipid droplet formation (see Introduction). Seipin has a highly conserved, central domain located in the ER lumen, and a divergent C-terminal tail that extends into the cytosol (Fig 14B). Both dominant and recessive mutations in the luminal domain and the C-terminal tail are responsible for severe human diseases [75]. seip-1(zu483) is a semi-dominant, compound mutation affecting the cytoplasmic tail; it causes an I260T substitution, and includes an additional frameshift mutation that changes several amino acids at the C-terminus (Fig 14B). COPA-1 is the sole C. elegans ortholog of α-COP, a COPI subunit. Recent genetic and biochemical studies have demonstrated a role for the COPI complex in regulating lipid droplets (see Introduction). However, the best known function of COPI vesicles is in the retrograde transport of ER-resident proteins from the Golgi to the ER (reviewed in [76]). ER-resident membrane proteins that are missorted into the secretory pathway, or that are temporarily shuttled to the Golgi for posttranslational modifications, can be returned to the ER through retrieval motifs recognized by α-COP and a related COPI subunit called β’-COP. The best characterized retrieval motifs for ER-resident membrane proteins are C-terminal dibasic peptides that are both necessary and sufficient for retrieval; the peptide motifs include KKxx, RKxx, and KxKxx, with the -3 lysine being the most critical residue. Structural studies have shown that an N-terminal WD40-repeat region in both α-COP and β’-COP forms a beta-propeller domain with a central binding pocket for the retrieval peptide [77,78]. The semi-dominant copa-1(zu482) mutation substitutes asparagine for a highly conserved, acidic residue in the binding pocket for the dibasic retrieval peptide (Fig 14C). Although retrieval sequences have not been analyzed in C. elegans, we found that several known and predicted ER-resident membrane proteins in C. elegans have conserved C-terminal sequences that match the KKxx, KxKxx or variant motifs (S10 Fig).

The nLD phenotype of COPI mutants results from a retrieval defect We wanted to determine whether the copa-1 mutation was affecting nLDs by a retrieval defect or through the other roles ascribed for COPI components in lipid droplets (see Introduction). Our genetic screens identified three additional, independent mutants with excess nLD phenotypes similar to copa-1(zu482). For example, the sizes and temperature dependence of nLDs in these mutants were similar to copa-1(zu482), but different from the nemp-1 and seip-1 mutants (S7 Fig). Because the COPI subunits α-COP and β’-COP have similar roles in binding retrieval peptides, we sequenced both copa-1/α-COP and copb-2/β’-COP in each mutant. None of the additional mutants had a mutation in copb-2, but one had the identical mutation as copa-1(zu482) (Fig 13C). We next asked whether copb-2 mutants with nLD phenotypes might have been recovered in the C. elegans "million mutation" project [79]. This study generated 2000 heavily mutagenized but viable strains that were grown clonally, sequenced, and then archived as a library of frozen stocks. Database searches showed that the library contained multiple strains with missense mutations in the WD40 repeat domains of either copa-1 or copb-2 [80]. None of strains with copa-1 mutations had nLD phenotypes, but two of 7 strains with copb-2 mutations (gk900572 and gk936767) had nLD phenotypes similar to copa-1(zu482). Both the copb-2(gk900572) and copb-2(gk936767) mutations altered amino acids close to critical residues in the predicted binding pocket for the retrieval motif (S10 Fig). Because the library strains have an average of about 400 mutations, we outcrossed each strain eight times to wild type by scoring for the nLD phenotype. DNA sequencing showed that both of the outcrossed lines retained their respective copb-2 mutations, indicating close linkage with the nLD phenotype. Finally, we used Crispr-Cas9 editing to generate a copb-2(zu521) mutation identical to copb-2(gk936767), and found that it resulted in a similar, semi-dominant nLD phenotype (Fig 13B and 13C).

Multiple ER-residents involved in fat regulation likely contribute to the nLD phenotype of COPI mutants The above results support a hypothesis that the nLD phenotypes of COPI mutants result from the failure to retrieve one or more ER-resident membrane protein(s). Interestingly, SEIP-1 is an ER-resident membrane protein, and the cytoplasmic domain of SEIP-1 shows little conservation in Caenorhabditis other than the C-terminal RKRK peptide (S10 Fig). RKRK is a functional retrieval peptide in other systems [81] and is absent in the seip-1(zu483) mutant (Fig 14B). We found that copa-1(zu482); seip-1(tm4221 null) double mutants showed a marked reduction, or complete absence, of germ cell nLDs, suggesting that SEIP-1 function contributes to the excess nLD phenotype of copa-1(zu482) single mutants (S9 Fig). We next used Crispr-Cas9 gene editing to change the presumptive retrieval peptide in SEIP-1 from RKRK to DLGS (Fig 14B). Although the resulting seip-1(syb2121) mutant had an excess nLD phenotype in germ cells, the phenotype was variable and minor compared with that of seip-1(zu483) or either COPI mutant (Fig 14D and 14E). We next used Crispr-Cas9 to engineer a complementary seip-1(zu520) mutation that retained the C-terminal RKRK peptide, but deleted several adjacent residues (Fig 14B). The seip-1(zu520) mutant had a strong, excess nLD phenotype resembling that of seip-1(zu483) (Fig 14D and 14E). Moreover, the nLDs in seip-1(zu520) were abnormally large (Fig 14D), with sizes similar to nLDs in seip-1(zu483) mutants, but much larger than nLDs in wild-type, seip-1(syb2121), copa-1(zu482), or copb-2(zu521) mutants. These results show that defects in the presumptive retrieval peptide of SEIP-1 contributes to the excess nLD phenotype of seip-1(zu483) mutants, but is not the sole or principal cause of the phenotype. To identify different, or additional, ER-resident membrane proteins that might contribute to the excess nLD phenotype of the COPI mutants, we searched a C. elegans database of predicted membrane proteins to find ER-resident proteins that (1) contained a C-terminal dibasic motif of the form KKxx or KxKxx, and (2) functioned in fat (triacylglyercol) synthesis [82]. This search identified ACL-2 and ACL-5 as potential candidates (Fig 14F). The major pathway that converts glycerol-3-phosphate to triacylglycerol involves the sequential addition of acyl groups from acyl-CoA to the glycerol backbone. The first acyl group is added by GPAT acyltransferases that are rate limiting for fat synthesis, and the second group is added by LPAAT acyltransferases. C. elegans has two non-mitochondrial GPATs, and two LPAATs [83]. However, only ACL-5/GPAT and ACL-2/LPAAT have candidate C-terminal retrieval peptides (Fig 14F). Both motifs are conserved in ACL-5 and ACL-2 proteins in Caenorhabditis species, suggesting that they have functional significance (Fig 14F). Thus, we used Crispr-Cas9 editing to change the critical -3 lysine in the dibasic motifs to either threonine [acl-5(zu525)] or to serine [acl-2(zu524)] (Fig 14F). The acl-5(zu525), but not acl-2(zu524), mutation caused a clear nLD phenotype in germ cells (Fig 14D, quantified in Fig 14G). Moreover, triple mutant seip-1(syb2121) acl-2(zu524); acl-5(zu525) animals had a much stronger nLD phenotype than the single or double mutants. These results support a hypothesis that germ cell nLDs can result from a defect in the COPI-mediated retrieval of ER-resident membrane proteins involved in fat synthesis. However, an important but unresolved issue is how COPI vesicles and retrieval peptides, which are thought to function in retrograde transport from the Golgi to the ER, prevent the accumulation of ER-residents in germ nuclei (see Discussion).

seip-1 mutant nLDs resemble wild type nLDs except for size and number We chose seip-1(zu483) mutants for a detailed characterization of nLDs and their effects on development because the nemp-1 and copa-1 mutants had additional phenotypes that were unlikely to be related directly to nLDs (see analysis in S11 Fig). For example, copa-1(zu482) males appeared to have a moderate mating or infertility defect, although nLDs are rare in males (S11 Fig). We found that nLDs in seip-1 mutant germ cells closely resembled those in wild-type germ cells, except for their size. The nLDs lacked any evidence of a heterochromatin coat, but many had "bristle coats" similar to some wild-type nLDs (Fig 11C). Germ nuclei with nLDs typically had only one nLD in seip-1(zu483) mutants (97%, n = 681), as observed for wild type but in contrast to copa-1(zu482) mutants (Fig 13A). Single nLDs could suggest that formation is triggered by a unique event in meiotic progression, or that an nLD occupies an exclusive site on the nuclear envelope. In our TEM analysis we looked for, but did not observe, unique features of the envelope near the early nLDs, including the position of the centrosome or a connection with the ER. Moreover, seip-1 mutant germ nuclei lacked the nuclear vesicles, membrane tubules, nMFBs, and cysts observed in wild-type intestinal nuclei. nLDs in seip-1 mutant gonads were not detected in the mitotic zone (n = 63 gonads) but were present in early pachytene cells (Fig 15A). This localization agrees with the earliest, albeit rare, examples of nLDs in wild type (Fig 10H). In comparing cells in the two regions by TEM, we noticed that interphase nuclei in the mitotic zone had abundant peripheral heterochromatin, similar to intestinal nuclei, but that early pachytene nuclei had very little peripheral heterochromatin (Fig 15B). In addition, early pachytene nuclei showed a marked increase in the numbers of ribosomes on the ONM (Fig 15B), suggesting that nLD formation might be linked to an increase in ONM-associated protein synthesis. Lamin appeared to be concentrated at the base of each of the nLDs in early pachytene nuclei (arrowheads in Fig 15C), and persisted through much of the pachytene stage. A similar concentration of lamin was not visible in any germ cells in the mitotic zone or in intestinal cells, and only very rare germ cell nLDs were completely surrounded by lamin (Fig 15D). Most of the nLDs in seip-1(zu483) germ cells remained adjacent to the envelope throughout pachytene, and the largest envelope-associated nLDs were associated with prominent nuclear bulges (arrows, Fig 11E). A subset of nLDs appeared to move away from the envelope and into the nucleolus, and those nuclei were spherical (panels 5, 6 in Fig 11E). PPT PowerPoint slide

PowerPoint slide PNG larger image

larger image TIFF original image Download: Fig 15. Characterization of nLDs in seip-1(zu483) gonads. (A) Mitotic to early pachytene region of a seip-1(zu483) gonad, showing the first appearance of nLDs (arrows and inset) in cells passing through the transition zone (TZ, see Fig 10A). (B) TEM of seip-1(zu483) germ cells. The left panel shows an interphase nucleus from the mitotic zone, which lacks nLDs. The right panel shows an early pachytene-stage nucleus with an nLD. The early pachytene nucleus has relatively little peripheral heterochromatin (hc), and far more ribosomes on the ONM (arrows). (C) Early pachytene region of seip-1(zu483) gonad showing lamin concentrated in patches at the base of each nLD. The inset is an orthogonal view of a single lamin patch, showing the asymmetric, horseshoe shape. (D) seip-1(zu483) germ cell showing a rare example of an nLD that appears to be coated with lamin (long arrow). The germ cell is also atypical in containing a second nLD (short arrow). (E) Image of the nLD-containing area of a seip-1(zu483) D2 gonad, from early pachytene to the loop region. The gonad is stained for lamin (white, LMN-1/lamin), lipid (green, BODIPY), and DNA (blue, DAPI). The projections are of 6 μm z-stacks. Essentially all lipid droplets visible in the nuclear level are nLDs, while the core projection shows cLDs in the core and nLDs in the flanking, peripheral nuclei. Note that the core cLDs are much smaller than the nLDs. The bottom panel shows nuclei (DAPI, blue) in a single focal plane through the middle of the gonad; the arrow at far left indicates transition zone nuclei as in Fig 15A. Arrowheads point to apoptotic nuclei with compacted DNA, and the double arrowhead indicates a binucleate apoptotic cell (see also Fig 11D). Note that large numbers of germ cells retain nLDs at they pass through the major death zone. (F) Comparison of nuclear sizes in the mid-pachytene to loop region of a D1 seip-1(zu483) gonad; staining shown for lamin (red, LMN-1), lipid (green, BODIPY), and DNA (blue, DAPI). The image is a single optical plane through the top nuclear level. Most nuclei increase in size uniformly from distal to proximal, but three nuclei (arrowheads) are noticeably smaller than their neighbors; the DNA panel shows that these nuclei are apoptotic with compacted chromatin. The large, lipid-containing objects that are not nuclear (asterisks) are within phagocytic sheath cells that engulf apoptotic germ cells (see S11 Fig for additional details). (G) Comparison of germ cell sizes in a seip-1(zu483) gonad; cell boundaries are outlined by F-actin (yellow, phalloidin), and nuclei are stained for lamin (red, LMN-1). Note the uniform increase in cell sizes as cells move distal to proximal; the insets show nLDs from two regions, including an nLD in an oocyte nucleus. (H) Binucleate cells in the early pachytene region of a seip-1(zu483) gonad; cell boundaries are indicated by staining for F-actin (yellow, phalloidin). The gonad contains low numbers of binucleate germ cells (T-bars), similar to wild type. Binucleate cells first appear in about the same region as nLDs, and undergo apoptosis as they move further proximal. When present, nLDs were usually found in both nuclei (23/28 binucleate cells), but could instead be in only one nucleus (5/28 binucleate cells). (I) Plot comparing the rate of larval development in apoptosis-defective ced-3(n717) mutants (n = 1581 larvae) with ced-3n717);seip-1(zu483) double mutants (n = 1445 larvae). Synchronous newly hatched larvae were allowed to develop and scored when they produced their first eggs. Most animals in both populations reached adulthood and began producing eggs at t = 49–50 hrs, but a small fraction of animals had not produced eggs by t = 60 hrs. https://doi.org/10.1371/journal.pgen.1009602.g015 nLDs in seip-1 mutant germ cells increased progressively in size during pachytene, similar to wild type (Fig 15E and 15F). The size of a nLD relative to the nuclear diameter in seip-1(zu483) germ cells could be as large or larger than in wild-type intestinal cells, but we did not find any examples of ruptured germ nuclei (n>2000 germ cells). Similar to wild type, the percentage of seip-1 mutant nuclei with nLDs appeared to decline after the death zone (Fig 15E and 15F). However, some nLDs in seip-1 mutant gonads persisted in enlarged oogonia, and a few mature, cellularized oocytes had nLDs (inset, Fig 15G).

[END]

[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1009602

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.