Differential impact of a dyskeratosis congenita mutation in TPP1 on mouse hematopoiesis and germline

Abstract

Telomerase extends chromosome ends in somatic and germline stem cells to ensure continued proliferation. Mutations in genes critical for telomerase function result in telomeropathies such as dyskeratosis congenita, frequently resulting in spontaneous bone marrow failure. A dyskeratosis congenita mutation in TPP1 (K170∆) that specifically compromises telomerase recruitment to telomeres is a valuable tool to evaluate telomerase-dependent telomere length maintenance in mice. We used CRISPR-Cas9 to generate a mouse knocked in for the equivalent of the TPP1 K170∆ mutation (TPP1 K82∆) and investigated both its hematopoietic and germline compartments in unprecedented detail. TPP1 K82∆ caused progressive telomere erosion with increasing generation number but did not induce steady-state hematopoietic defects. Strikingly, K82∆ caused mouse infertility, consistent with gross morphological defects in the testis and sperm, the appearance of dysfunctional seminiferous tubules, and a decrease in germ cells. Intriguingly, both TPP1 K82∆ mice and previously characterized telomerase knockout mice show no spontaneous bone marrow failure but rather succumb to infertility at steady-state. We speculate that telomere length maintenance contributes differently to the evolutionary fitness of humans and mice.

Figure 1.. CRISPR-Cas9 generation of a mouse model of a dyskeratosis congenita mutation in TPP1.

(A) Overlay of the crystal structures of human TPP1 OB (hTPP1 OB) domains from WT and K170∆ proteins. The deletion of K170 distorts the acidic TEL patch knuckle. (B) Strict conservation of the TEL patch loop (acidic amino acids shown in red) that also harbors the human TPP1 K170 residue (shown in cyan) that is deleted in dyskeratosis congenita. The mouse equivalent of this residue is K82. (C) The PAM sites (GGG and AGG highlighted in red and orange, respectively) and Cas9 cleavage sites (arrowheads) for the two guide RNAs used to cleave the exon coding for mouse TPP1 K82∆ are shown. The schematic for the homologous recombination repair template shows not only the deletion of the K82 codon (red) but also a silent mutation that creates a KpnI site (green) for screening purposes. (D) PCR amplification and KpnI restriction digestion screening of eight blastocysts after injection with guide RNAs and repair template for introducing the TPP1 K82∆ mutation. Expected sizes for PCR amplicon and its KpnI cleavage products used to screen the editing of the Acd locus of mouse blastocysts by CRISPR-Cas9 are shown above the gel. (D, E) Sanger sequencing of the PCR products of the indicated blastocysts (same as those analyzed in panel D) showing accurate editing of the Acd locus. (F) Images of G1 WT and K82∆ (homozygous), male and female mice. (G) Breeding scheme to backcross the CRISPR-edited K82∆/+ founder mouse and generate WT and homozygous K82∆ mice that were bred for five generations (G1 → G5).

Figure S1.. Establishment of Flow-FISH-based approach to determine BM telomere length of WT and K82∆ mice.

(A) Schematic of Flow-FISH procedure optimized for mouse BM. (B) Each mouse sample is run separately (visualized as grey DNA strand) and in combination with calf thymocytes (visualized as purple DNA strand; used as an internal control for telomere length and as a calibration control for determining the absolute length of mouse telomeres). Each mouse replicate includes a non-probe control to account for any autofluorescence. Red DNA strand indicates DNA with telomeric PNA probe. (C) Telomere restriction fragment Southern blot of genomic DNA from calf thymocytes and TPP1-S stable cell line (Grill et al, 2019) with two concentrations of λ HindIII digest DNA ladder. (C, D, E, F, G, H, I) Representative flow cytometry plots of single controls used (D, G) calf thymocytes-only, (E, H) BM-only, and (F, I) combined. (D, E, F) Gates were drawn based on nucleic stain signal on the y-axis and forward scatter on the x-axis and because of differences in size can be gated separately. (G, H, I) Better distinction between populations can be achieved when gated with nucleic stain signal on the y-axis and telomeric DNA probe on the x-axis instead of FSC-H. (J) Cumulative data for the quantitation of telomere lengths across generations showing decreased telomere length in each generation of K82∆ mice. G4 WT telomere length was not determined (n.d.). No sex differences were observed.

Figure 2.. Flow-FISH to measure telomere length reveals progressive telomere shortening in K82∆ mutant mice.

(A, B) Representative flow cytometry plots of Flow-FISH of WT G1 mouse samples with (A) no probe and (B) PNA telomere probe. Gating includes calf thymocytes (+/− probe) and BM (+/− probe). (C) The histogram shows data with the probe. Red indicates calf thymocyte telomeric probe signal. Grey indicates BM telomeric probe signal. (D) Histogram showing telomeric probe signal of G1 WT (grey filled peak) and K82∆ (grey open peak) alongside their internal calf thymocyte controls used in the same experiment. (E) Quantitation of absolute telomere length of G1 WT and K82∆ calibrated against calf thymocytes telomere length obtained by TRF analysis. (F) Histogram showing telomeric probe signal of G5 WT (black filled peak) and K82∆ (black open peak) alongside their calf thymocyte controls used in the same experiment. (G) Quantitation of absolute telomere length of G5 WT and K82∆ calibrated against calf thymocytes telomere length obtained by TRF analysis. n = 4–6 mice per condition; mean with 95% CI; significance calculated with Prism software using t test for individual experiments; ***P ≤ 0.002, ****P ≤ 0.0001.

Figure S2.. Analysis of complete blood counts in K82∆ mice does not reveal defects in steady-state hematopoiesis.

(A, B, C, D, E) Complete blood counts at a given generation and age indicated on the x-axis showing no consistent defect in hematopoiesis. Red circles indicate female WT mice, black circles indicate male WT mice, and corresponding blue circles indicate mutant K82∆ mice. (A, B, C, D, E) Specific parameters studied include (A) white blood cells, (B) platelets, (C) red blood cells, (D) hemoglobin, and (E) mean corpuscular volume. G1 (16 mo) WT: n = 18 (9 males and 9 females); G1 (16 mo) K82∆: n = 19 (9 males and 10 females). G1 (2 yr) WT: n = 9 (6 males and 3 females); K82∆: n = 11 (6 males and 5 females). G3 (7 mo) WT: n = 24 (12 males and 12 females); K82∆: n = 24 (12 males and 12 females). G3 (12 mo) WT: n = 19 (8 males and 11 females); K82∆: n = 20 (10 males and 10 females). G3 (16 mo) WT: n = 18 (7 males and 11 females); K82∆: n = 20 (9 males and 11 females). Mean with 95% confidence interval. * <0.05, ** <0.01, *** <0.001 for P-values determined by t test.

Figure S3.. Analysis of mature and progenitor BM cells in K82∆ mice does not reveal defects in hematopoiesis.

(A, B, C, D, E, F, G, H, I, J, K, L, M, N, O) Quantitation across G3-G5 mice; red filled circles indicate female WT mice, red open circles indicate female K82∆ mice, black filled circles indicate male WT mice, and black open circles indicate male K82∆ mice. (A, B, C, D, E, F, G, H, I, J, K, L, M, N, O) Specific parameters studied include (A) cellularity, (B) CD11b⁺ GR-1⁺ myeloid cells, (C) Ter119⁺ erythroid cells, (D) B220⁺ CD43⁻ sIgM⁻ CD93⁺ Pre B cells, (E) B220⁺ CD43⁺ CD19⁺ CD93⁺ Pro B cells, (F) CD19⁺ B220⁺ B cells (G) B220⁺ CD43⁻ sIgM⁺ CD93⁺ immature B cells, (H) Lin⁻ Sca-1⁻ cKIT⁺LK cells, (I) Lin⁻ Sca-1⁺ cKIT⁺ LSK cells, (J) CD150⁺ CD48⁻ LSK cells (LT-HSCs), (K) Lin⁻ Sca-1⁻ cKIT⁺ CD150⁺ CD41⁺ MkP cells, (L) Lin⁻ Sca-1⁻ cKIT⁺ CD41⁻ CD16-32⁺ CD150⁻ GMP cells, (M) Lin⁻ Sca-1⁻ cKIT⁺ CD41⁻ CD16-32⁻ CD150⁻ CD105⁻ Pre GM cells, (N) Lin⁻ Sca-1⁻ cKIT⁺ CD41⁻ CD16-32⁻ CD150⁺ CD105⁻ Pre MegE cells, (O) Lin⁻ Sca-1⁻ cKIT⁺ CD41⁻ CD16-32⁻ CD150⁺ CD105⁺ Pre CFU-E cells. G3 WT: n = 18 (7 males and 11 females); K82∆: n = 20 (9 males and 11 females). Mean with 95% confidence interval. * <0.05, ** <0.01; ^M or ^F indicates significance for male (M) or female (F). See the Materials and Methods section for more information about antibodies (manufacturer, clone names).

Figure 3.. K82∆ mutant mice do not develop BM failure.

(A, B) BM cellularity for 1 hindlimb per mouse for WT and K82∆ mice in (A) G4 and (B) G5. (C, D, E, F, G, H, I, J, K) Representative (C, D) lineage negative (Lin⁻), (F, G) Lin⁻Sca-1⁺c-Kit^high (LSK), and (I, J) CD150⁺ CD48⁻ LSK cells (LT-HSCs) flow cytometry plots of G5 WT (C, F, I), G5 K82∆ mice (D, G, J) with quantitation (E, H, K) showing equal frequencies in K82∆ mice compared with WT. Red indicates female mice, black indicates male mice, filled symbols indicate WT, and open symbols indicate K82∆ mice. n ≥ 4; mean with 95% confidence interval.

Figure S4.. Analysis of splenocytes in K82∆ mice does not reveal defects in spleen-resident populations.

(A, B, C, D, E, F, G) Quantitation across G3-G5 mice; red filled circles indicate female WT mice, red open circles indicate female K82∆ mice, black filled circles indicate male WT mice, black open circles indicate male K82∆ mice. (A, B, C, D, E, F, G) Specific parameters studied include (A) cellularity, (B) CD11b⁺ GR-1⁺ myeloid cells, (C) CD19⁺ B220⁺ B cells, (D) CD8⁺ T-cells, (E) CD4⁺ T-cells, (F) TCRB⁺ T-cells, and (G) Ter119⁺ erythroid cells. G4 WT: n = 20 (11 males and 9 females); K82∆: n = 22 (11 males and 11 females). Mean with 95% confidence interval. * <0.05, ** <0.01; ^M or ^F indicates significance for male (M) or female (F). See the Materials and Methods section for more details on antibodies used.

Figure S5.. Analysis of thymocytes in K82∆ mice does not reveal defects in T lineage development.

(A, B, C, D, E, F, G) Quantitation across G3-G5 mice; red filled circles indicate female WT mice, red open circles indicate female K82∆ mice, black filled circles indicate male WT mice, and black open circles indicate male K82∆ mice. (A, B, C, D, E, F, G) Specific parameters studied include (A) cellularity, (B) Lin⁻ CD8⁻ double-negative cells, (C) CD44⁺ c-KIT⁺ Lin⁻ CD8⁻ double-negative 1 cells, (D) CD25⁻ CD44⁺ c-KIT⁺ Lin⁻ CD8⁻ early thymic progenitor cells, (E) CD25⁺ CD44⁺ c-KIT⁺ Lin⁻ CD8⁻ double-negative two cells, (F) c-KIT^low CD25⁺ Lin⁻ CD8⁻ double-negative three cells, and (G) CD8⁺ CD4⁺ double-positive cells. G5 WT: n = 9 (4 males and 5 females); K82∆: n = 9 (4 males and 5 females). Mean with 95% confidence interval. * <0.05, ** <0.01, *** <0.001; ^M or ^F indicates significance for male (M) or female (F). See the Materials and Methods section for more details on antibodies used.

Figure 4.. TPP1 K82∆ mutation leads to reproductive defects in later generations.

(A, B) X-axis indicates genotype of the parents. (A) Quantitation of the number of offspring per litter for the indicated generation number and genotype analyzed in this study. (B) Quantitation of the total number of pups that survived past weaning per mating pair for the indicated generation number of WT and K82∆ mice. Number of mating pairs used for data in panels (A) and (B): G1 WT: n = 4; G1 K82∆: n = 3; G2 WT: n = 8; G2 K82∆: n = 10; G3 WT: n = 9; G3 K82∆: n = 3; G4 WT: n = 2; G4: K82∆ n = 13; G5 WT: n = 3; G5 K82∆: n = 4. (C) Representative gross morphology images of testes for the indicated generation number and genotype of mice studied. (D) Quantitation of testes/body weight of WT and K82∆ mice. n = 4–11. (E, F) Whole testis sections were stained with H&E and imaged and analyzed for seminiferous tubule (E) diameter and (F) number in G1 and G5, WT, and K82∆, mice. At least two mice were evaluated per generation and genotype. Slides separated by at least 100 μm were used as technical replicates. (G) Number of sperm collected from vas deferens and epididymis from G5 WT and G5 K82∆ mice. n = 4. P = 0.0067. (H) Representative sperm morphology from G5 WT and G5 K82∆ mice by H&E staining. Magnified views of boxed areas are shown on the right. Scale bar: 20 μm. (I) Quantitation of the percentage of abnormal sperm. n = 4.

Figure S6.. Schematic for seminiferous tubule organization and spermatogenesis.

(A) Cartoon of testis created on biorender.com. (B) Magnified view of a seminiferous tubule with a cross section of germ cells and somatic Sertoli cells. (C) Cartoon of a seminiferous tubule cross section showing Sertoli cells (green), spermatogonia (dark purple), primary spermatocytes (light purple), secondary spermatocytes (pink), and spermatids/spermatozoa in the lumen. (D) Schematic to show the spatial orientation of germline cells within a tubule. (E) Representative image of a G5 WT testis cross section. Sertoli cells (green), spermatocytes (red/purple), nuclei (blue), spermatids (grey). Scale bar: 50 μm.

Figure 5.. Late-generation K82∆ mice testes are comprised primarily of disordered or empty tubules.

(A) Immunofluorescence (IF) for PNA-lectin (spermatid acrosomes; grey) and DAPI (nuclei; blue) in cross sections of testes from G5 WT and G5 K82∆ mice in each stage including those that were not-stageable (N.S.) because they were devoid of spermatids, which are necessary for accurately staging tubules. Scale bar: 50 μm. (B, C) Quantitation of total tubules that are stageable or not-stageable in (B) G1 WT, G1 K82∆ mice, and acd/acd or (C) G5 WT, G5 K82∆, and acd/acd mice. (D, E) Quantitation of the percentage of stageable seminiferous tubules in a given stage of spermatogenesis in (D) G1 WT and G1 K82∆ mice and (E) G5 WT and G5 K82∆ mice. (F) Quantitation of the breakdown of each stage within G1 WT, G1 K82∆, G5 WT, G5 K82∆, and acd/acd mice. White diagonal lines indicate % of tubules in stages I–III, black horizontal lines are stages IV–VI, grey vertical lines are stages VII–VIII, black diagonal lines are stages IX–XII, and black filled bars are not-stageable. (G) Quantitation of tubule phenotype denoted as either normal organization (white bar), disordered organization (grey bar), or empty tubules (black bar) in G1 WT, G1 K82∆, G5 WT, G5 K82∆, and acd/acd mice. (H, I) Quantitation of the percentage of seminiferous tubules with either normal, disordered, or empty tubule phenotype in (H) G1 WT, G1 K82∆, and acd/acd mice and (I) G5 WT, G5 K82∆, and acd/acd mice. acd/acd was used as a positive control of gonadal defect. In panels (B, D, H): G1 WT (grey filled circles), G1 K82∆ (grey open circles). In panels (C, E, I): G5 WT (black filled squares), G5 K82∆ (black open squares). In panels (B, C, H, I): acd/acd (red open triangle).

Figure 6.. K82∆ mutation results in a reduction of germ cells but an increase in somatic Sertoli cells.

(A, C, E) Immunofluorescence for (A) PLZF (undifferentiated spermatogonia; red) (C) SCP3 (spermatocytes; red), (E) SOX9 (Sertoli cells; red), DAPI (nuclei; blue), and PNA Lectin (spermatid acrosomes; grey) in cross sections of testes from G5 WT, G5 K82∆ mice, and acd/acd mice. Scale bar: 50 μm. (B, D, F) Quantitation of the number of (B) spermatogonia, (D) spermatocytes, and (F) Sertoli cells per tubule in G1 WT, G1 K82∆ G5 WT, G5 K82∆, and acd/acd mice. See the Materials and Methods section for total number of mice and tubules analyzed.

Figure S7.. K82∆ mutant testes do not show changes in proliferation or apoptosis compared with WT mice.

(A, B) Immunofluorescence for (A) BrdU (proliferation; green) or (B) cleaved Caspase-3 (apoptosis; green); DAPI (nuclei; blue), and PNA Lectin (spermatid acrosomes; grey) in cross sections of testes from G5 WT and G5 K82∆ mice. Scale bar: 50 μm. (C, D) Quantitation of the percentage of (C) BrdU-positive and (D) cleaved Caspase-3-positive cells per tubule in G1 WT, G1 K82∆, G5 WT, and G5 K82∆ mice.