A PUF Hub Drives Self-Renewal in Caenorhabditis elegans Germline Stem Cells

Abstract

Stem cell regulation relies on extrinsic signaling from a niche plus intrinsic factors that respond and drive self-renewal within stem cells. A priori, loss of niche signaling and loss of the intrinsic self-renewal factors might be expected to have equivalent stem cell defects. Yet this simple prediction has not been borne out for most stem cells, including Caenorhabditis elegans germline stem cells (GSCs). The central regulators of C. elegans GSCs include extrinsically acting GLP-1/Notch signaling from the niche; intrinsically acting RNA-binding proteins in the PUF family, termed FBF-1 and FBF-2 (collectively FBF); and intrinsically acting PUF partner proteins that are direct Notch targets. Abrogation of either GLP-1/Notch signaling or its targets yields an earlier and more severe GSC defect than loss of FBF-1 and FBF-2, suggesting that additional intrinsic regulators must exist. Here, we report that those missing regulators are two additional PUF proteins, PUF-3 and PUF-11 Remarkably, an fbf-1fbf-2 ; puf-3puf-11 quadruple null mutant has a GSC defect virtually identical to that of a glp-1/Notch null mutant. PUF-3 and PUF-11 both affect GSC maintenance, both are expressed in GSCs, and epistasis experiments place them at the same position as FBF within the network. Therefore, action of PUF-3 and PUF-11 explains the milder GSC defect in fbf-1fbf-2 mutants. We conclude that a "PUF hub," comprising four PUF proteins and two PUF partners, constitutes the intrinsic self-renewal node of the C. elegans GSC RNA regulatory network. Discovery of this hub underscores the significance of PUF RNA-binding proteins as key regulators of stem cell maintenance.

Keywords: FBF; PUF-11; PUF-3; redundancy; stem cell regulatory network.

Figure 1

puf-3 and puf-11 are putative missing GSC regulators. (A) Diagram illustrating functional relationships between key genes that regulate stem cell self-renewal and differentiation in C. elegans germline stem cells (GSCs) (Kimble and Crittenden 2007; Kershner et al. 2014). Gene X (red) represents a missing factor that likely functions in parallel to fbf-1 and fbf-2. (B) Total number of germ cells (GCs) per animal at 20° in wild type (wt) and strong loss-of-function mutants of GSC regulators: fbf-1 fbf-2, lst-1 sygl-1, and glp-1; data adapted from the literature (Austin and Kimble 1987; Crittenden et al. 2002, 2006; Kimble and Crittenden 2005; Kershner et al. 2014). x-axis: L1–L4, larval stages of development; ticks mark molts between stages. (C) Total GC number produced per animal; data generated in this work. Animals of each strain were staged to 36 hr (15°), 24 hr (20°), or 18 hr (25°) past mid-L4 for this experiment. In germlines where distal germ cells had differentiated to mature sperm, GC number was determined by counting sperm number and dividing by four. GC number were not counted in larger germlines that retained a PZ; these were scored as “many and still dividing.” n indicates the number animals scored. Alleles used here and throughout this work are strong loss of function and likely null: glp-1(q46), lst-1(q869), sygl-1(q828), fbf-1(ok91), and fbf-2(q704). (D) Schematic of puf-3 and puf-11 loci, which share 90% nucleotide (nt) identity (Hubstenberger et al. 2012): exons (white boxes), introns (peaked lines), untranslated region (gray boxes), sequence encoding individual PUF repeats (blue circles), and sequence targeted by RNAi clone (orange line). (E) Diagram of distal gonad with the PZ and germ cells that have entered early meiotic prophase (Diff). Distal end (asterisk); GSCs (yellow); GSC daughters primed for differentiation and transitioning toward entry into meiotic prophase, including those in meiotic S phase (graded yellow to green); and early meiotic prophase (green, crescent-shaped). (F and G) Representative images of extruded gonads from 25° fbf-1 fbf-2 adult hermaphrodites staged to 18 hr after L4. RNAi treatment, either empty vector control (F) or puf-3/11 RNAi clone B (G), was initially delivered to these animal’s mothers and then continued from hatching, through larval development and into adulthood. Gonads are immunostained with a sperm marker α-SP56 (red) and DAPI (cyan). Images are Z-projections of several fluorescent images obtained on a compound microscope. A dotted line delineates the gonad boundary and an asterisk marks the distal end. Double-headed arrow (yellow) indicates the PZ. Scale bar in F applies to both images. (H) State of distal germ cells of fbf-1 fbf-2 hermaphrodites and fbf-1 fbf-2; him-5 males, raised at 25° on either empty or puf-3/11 RNAi, and staged to adulthood (18 hr after L4). Distal germ cell states were determined by DAPI-stained chromosomal morphology, scoring for a PZ or mature sperm in the distal germline. him-5(e1490) increased the frequency of male progeny (Hodgkin et al. 1979). (I) Quantitation of total GC per animal from experiment in H by counting sperm number and dividing by four. In germlines with a PZ, GC numbers were not counted but appeared comparable to previous reports of >100 (Merritt and Seydoux 2010; Shin et al. 2017). RNAi treatment on x-axis: empty vector (e), puf-3/11 RNAi clone A (indicated by A), or puf-3/11 RNAi clone B (indicated by B). Individual data points are plotted as circles; middle line is median value. * marks an animal with a total of 47 GCs.

Figure 2

GSC maintenance defects are undetectable in single and double puf-3 and puf-11 mutants. (A) puf-3 and puf-11 loci, using conventions as in Figure 1D. Extents of deletion mutants are bracketed below loci and position of the point mutation is marked with an asterisk above. See Figure S2 for sequence details. (B) Germline-related characteristics of single and double puf-3 and puf-11 mutants at 20°. See Materials and Methods for details about assays and scoring. (C and D) Representative confocal Z-projections of DAPI-stained gonads extruded from animals staged to 24 hr after L4 at 20°. Extent of PZ, double-headed yellow arrow. Annotation by convention in Figure 1, F and G and scale bar in C applies to D. (E and F) PZ sizes measured in number of germ cells (GCs). (F) PZ sizes showing individual data points as circles; middle line, median; boxes, 25–75% quantile; whiskers, minimum and maximum values. n = 5 gonad arms for each sample. P-values: mutants were compared to wild type using Welch’s ANOVA and Games–Howell post hoc test.

Figure 3

Triple and quadruple mutants reveal puf-3 and puf-11 role in GSC maintenance. (A) Total germ cell (GC) number per animal in fbf-1 fbf-2; puf triple mutant adults at 20° (left) and 25° (right). GC number was counted only in germlines differentiated to the distal end; germlines retaining a PZ or having meiotic prophase nuclear morphology were scored as “many.” Where distal germ cells had differentiated to mature sperm, GC number was determined by counting sperm and then dividing by four. Individual data points are plotted as circles; middle line, median; boxes, 25–75% quantile; whiskers, minimum and maximum values falling outside the box but within 1.5 times the interquartile range. To determine P-values, triple mutants were compared to fbf-1 fbf-2 control using Welch’s ANOVA and Games–Howell post-hoc test; n.s., not significant; ** P < 0.01, *** P < 0.001; n/a, not applicable. For additional information including mean GC/animal, see Figure S3. (B) State of distal germ cells in fbf-1 fbf-2; puf triple mutants at 25°, assayed in adults at 18 hr past L4. State was scored as described in Figure 1H, with the additional classification of “meiotic” for cells with meiotic prophase chromosomal morphology, but not yet differentiated as sperm. (C) Graph of germ cell number per animal in fbf-1 fbf-2; puf-3 puf-11 adults at 15, 20 and 25°. Germline scoring and graph conventions as in A. P-value compared to respective fbf-1 fbf-2 control was determined using Welch’s ANOVA and Games–Howell post-hoc test; ***P < 0.001; n/a, not applicable. For additional information including mean GC per animal, see Figure S3. (D and E) Representative images of DAPI-stained gonads extruded from wild type (D) or puf-3(q966) puf-11(q971) (E) adult hermaphrodites treated with fbf-1/2 RNAi at 25°. Animals were plated to RNAi as mid-L4s and analyzed for changes in nuclear morphology 48 hr later, when the RNAi treatment had had time to work effectively and animals had reached adulthood. The extent of the PZ (double-headed arrow, yellow) and “crescent-shaped nuclei” typical of early meiotic prophase (green) are annotated. Other annotation by convention in Figure 1, F and G. Images show single confocal Z-sections and scale bar in D applies to both images.

Figure 4

GSC self-renewal fails during larval development in fbf-1 fbf-2; puf-3 puf-11 quadruple mutants at 20°. (A–D) Representative images of early L4 germlines stained with α-PGL-1 (white), a germ cell marker. Images were obtained by fluorescence compound microscopy. The scale bar in A applies to all images. All other annotation by convention in Figure 1, F and G. (E and F) Quantitation of the number of germ cells (GCs) per animal across larval stages. Larvae were immunostained (as in A–D) at defined developmental timepoints: early L1 (eL1), early L2 (eL2), late L2, and early L4 (eL4) stages. GC number was counted as number of PGL-1–positive cells for all timepoints, except early L1, which was scored live by DIC microscopy. (E) Bars show the mean number of germ cells per sample with genotypes color coded as shown. Each individual data point is plotted as a gray circle. not done (nd): germ cells in glp-1 animals at early L4 stage had already differentiated to sperm and were thus not scored; the striped bar in early L4 is replicated from the late L2 mean value for glp-1. (F) Table with mean number of GCs per animal. (A–F) Alleles are as follows: fbf-1(ok91), fbf-2(q704), puf-3(q966), puf-11(q971), glp-1(q46).

Figure 5

PUF-3 and PUF-11 are expressed in distal germline, including GSCs. (A) Schematic of puf-3 and puf-11 loci with epitope tags annotated by convention as in Figure 1D. Inverted triangles denote insertion sites of 3xV5 epitope tags, which are flanked both up- and downstream by a GS linker. See Figure S2 for a more detailed sequence annotation. (B–G) Representative images of PUF-3^V5 and PUF-11^V5 expression in gonads extruded from L4 hermaphrodites raised at 20°. Gonads from puf-3^V5 [puf-3(q1058)] (B and E), puf-11^V5 [puf-11(q1128)] (C and F), and wild-type control (D and G) were stained with α-V5 (magenta) and DAPI (cyan) and then imaged by confocal microscopy. (B–D) are Z-projections. V5 signal intensity (int.) was adjusted uniformly in Adobe Photoshop across images, and high or low intensities are indicated on the left. In all genotypes, including the wild-type control lacking a V5 epitope, nonspecific staining was detected in the somatic distal tip cell with the α-V5 antibody. In E–G, a middle Z-slice from the respective gonad in B–D is shown. Magnified insets are outlined in yellow and labeled left (L) or right (R) corresponding to their position in the image; white arrowheads mark representative cytoplasmic granules. Scale bar in B applies to all images except the insets, where Bar, 2 μm. Other annotation conventions as in Figure 1, F and G. (H and I) Quantitation of V5 intensity in the distal region of PUF-3^V5 (light pink), PUF-11^V5 (dark pink), and wild-type (gray) extruded gonads, determined using Fiji/ImageJ (see Materials and Methods for details). Each bar represents the mean α-V5 immunostaining intensity in arbitrary units (a.u.) in the distal-most 50 μm of the gonad [∼11 germ cell diameters (gcd) using the conversion 4.4 μm/gcd (Lee et al. 2016)], with nonspecific staining background from the wild-type control subtracted. Error bars represent SE. (H) Analysis of mid-L4 staged extruded germlines at 20°. Four independent experiments were performed for a total of at least 27 gonads per genotype. (I) Analysis of germlines extruded from staged adults at 20° (24 hr past mid-L4) and 25° (18 hr past mid-L4). Two independent experiments were performed for a total of 24 gonads per experimental condition. For representative images at 20°, see Figure S4, E–G.

Figure 6

puf-3 puf-11 lie parallel to fbf-1 fbf-2 in the GSC regulatory pathway. (A) Results of epistasis tests conducted with lst-1(gf) and sygl-1(gf) alleles at 25°. Number of germ cells (GCs) in Glp animals was determined by counting sperm in adults and dividing by four. Genotype for lst-1(gf) strain is lst-1(ok814); qSi267 [Pmex-5::LST-1::3xFLAG::tbb-2 3′ end] fbf-1(ok91) fbf-2(q704) and sygl-1(gf) is sygl-1(tm5040); qSi235 [Pmex-5::SYGL-1::3xFLAG::tbb-2 3′ end] fbf-1(ok91) fbf-2(q704). We used the puf-3/11 RNAi clone B in these experiments. Images of representative germlines are available in Figure S5, A–D. (B) Results of epistasis tests conducted with gld-1 gld-2 at 20°. Genotype is gld-2(q497) gld-1(q361); fbf-1(ok91) fbf-2(q704). We used the puf-3/11 RNAi clone B in these experiments. Images of representative germlines are available in Figure S5, E and F. (C) Revised pathway for GSC regulation that includes puf-3 and puf-11 at the same position in the pathway as fbf-1 and fbf-2 (blue shading).

Figure 7

PUFs and FBF comprise a PUF hub that accounts for GSC self-renewal. (A) Yeast two-hybrid schematic. SYGL-1 was fused to the Gal4 activation domain (AD), which was HA tagged. PUF protein variants were fused to the LexA binding domain (BD), which was V5 tagged. PUF constructs included the PUF repeats and some flanking amino acids; for amino acid boundaries, see Materials and Methods. Interaction between SYGL-1 and PUF drives transcription of a lacZ (β-gal) reporter. (B) SYGL-1/PUF interaction was measured using β-galactosidase (β-gal) activity. Each bar is the mean of at least three independent experiments. Individual data points are plotted as gray circles. (C) Western blot from yeast lysate probed with α-HA to detect AD fusion proteins and α-V5 to detect BD fusion proteins. For the blots detecting BD::V5::PUF fusions, the molecular weight ladder is annotated at left and a red asterisk marks the relevant full-length protein in each lane. Expected sizes are as follows: FBF-1, 80.9 kDa; FBF-2, 82.4 kDa; PUF-3, 71.9 kDa; PUF-9, 85.9 kDa; and PUF-11, 71.7 kDa. Actin was the loading control. (D) Total number of germ cells (GCs) per animal at 20° in mutants of key GSC regulators, modified from Figure 1B to include fbf-1 fbf-2; puf-3 puf-11 (red triangles) (this work). (E) The PUF hub for GSC self-renewal consists of two PUF partners that are direct targets of niche signaling and four PUF RNA-binding proteins that collectively regulate a battery of messenger RNAs. (F) Molecular model for GSC self-renewal: PUF RNA-binding protein binds to the 3′ UTR of target messenger RNAs. A PUF partner, LST-1 or SYGL-1, ensures RNA repression by an unknown mechanism.