RpL22e, but not RpL22e-like-PA, is SUMOylated and localizes to the nucleoplasm of Drosophila meiotic spermatocytes

Abstract

Duplicated ribosomal protein (Rp) gene families often encode highly similar or identical proteins with redundant or unique roles. Eukaryotic-specific paralogues RpL22e and RpL22e-like-PA are structurally divergent within the N terminus and differentially expressed, suggesting tissue-specific functions. We previously identified RpL22e-like-PA as a testis Rp. Strikingly, RpL22e is detected in immunoblots at its expected molecular mass (m) of 33 kD and at increasing m of ~43-55 kD, suggesting RpL22e post-translational modification (PTM). Numerous PTMs, including N-terminal SUMOylation, are predicted computationally. Based on S2 cell co-immunoprecipitations, bacterial-based SUMOylation assays and in vivo germline-specific RNAi depletion of SUMO, we conclude that RpL22e is a SUMO substrate. Testis-specific PTMs are evident, including a phosphorylated version of SUMOylated RpL22e identified by in vitro phosphatase experiments. In ribosomal profiles from S2 cells, only unconjugated RpL22e co-sediments with active ribosomes, supporting an extra-translational role for SUMOylated RpL22e. Ectopic expression of an RpL22e N-terminal deletion (lacking SUMO motifs) shows that truncated RpL22e co-sediments with polysomes, implying that RpL22e SUMOylation is dispensable for ribosome biogenesis and function. In mitotic germ cells, both paralogues localize within the cytoplasm and nucleolus. However, within meiotic cells, phase contrast microscopy and co-immunohistochemical analysis with nucleolar markers nucleostemin1 and fibrillarin reveals diffuse nucleoplasmic, but not nucleolar RpL22e localization that transitions to a punctate pattern as meiotic cells mature, suggesting an RpL22e role outside of translation. Germline-specific knockdown of SUMO shows that RpL22e nucleoplasmic distribution is sensitive to SUMO levels, as immunostaining becomes more dispersed. Overall, these data suggest distinct male germline roles for RpL22e and RpL22e-like-PA.

Keywords: Drosophila; RpL22e; RpL22e-like-PA; SUMOylation; duplicated ribosomal proteins; male germline; phosphorylation; post-translational modification; ribosomal protein paralogues.

Figure 1. RpL22e is detected at higher m in multiple Drosophila tissues. (A) Using a peptide-derived polyclonal antibody, RpL22e is detected in S2 tissue culture cells, testis and ovary tissue at its predicted m of 33 kD, but also at increasing m, designated as 43α,β and 55α,β,γ (increasing m of ~43 kD and ~55 kD). (B) Peptide inhibition experiments confirm specificity of polyclonal antibody. (C) Immunodetection of 33 kD RpL22e, as well as novel slower migrating species (43α,β and 55α) is reduced in S2 cells via RNAi by incubation of dsRNA targeting codons 1–100 of RpL22e, but not by targeting GFP (negative control). Tubulin was used as a loading control.

Figure 2. Computational predictions of post-translational modifications within the RpL22e family. Eukaryotic Linear Motif (ELM) scanner predicts multiple modifications in both RpL22e (FBgn0015288; FBpp0070143) and RpL22e-like-PA (FBgn0034837; FBpp0071958) as consensus sequences were conserved for various phosphorylation and SUMOylation motifs. The black domains represent the conserved region between the fly paralogues and other eukaryotic RpL22e proteins. Dark and light gray domains represent the fly-specific histone H1-like N-terminal extension that has less homology between the paralogues. Consensus sequences and motifs within RpL22e family members are found in Table S1.

Figure 3. Anti-SUMO immunoprecipitates FLAG-tagged RpL22e of higher m. (A) FLAG-RpL22e is detected (using anti-FLAG Ab) above its predicted m of 33 kD in the 43–55 kD range in the 529SU S2 cell line, which harbors inducible expression of the SUMO protein (as FLAG-SUMO) and the E2 SUMO conjugating enzyme Ubc9 (as HA-Ubc9). Although the FLAG Ab will detect both FLAG-RpL22e and all SUMOylated proteins, we note that FLAG-SUMO conjugates are only detected when FLAG-RpL22e, but not GFP, is transiently expressed. Cells treated with calcium phosphate alone served as the mock control. The FLAG-SUMO conjugates detected are not present in the mock control. Tubulin was used as a loading control. (B) 55α RpL22e can be immunoprecipitated from S2 cells using anti-Drosophila SUMO, but not with beads alone. The polyclonal Ab against RpL22e was used for detection.

Figure 4. FLAG-RpL22e, but not FLAG-RpL22e-like-PA, can be SUMOylated in vitro. (A) When co-expressed in E. coli harboring the Drosophila SUMOylation machinery (the E1 heteromeric activating enzyme and E2 conjugating enzyme) with an attachable competent SUMO protein (Q^SUMO), but not with an incompetent SUMO mutant (Q^ΔGG), FLAG-RpL22e is detected above its predicted m (33 kD) at 55 kD (arrowhead). Based on the m shift, the addition of two SUMO moieties is predicted. (B) Although harboring two predicted SUMO motifs (Fig. 2, Table S1), FLAG-RpL22e-like-PA is not SUMOylated in vitro.

Figure 5. Testis RpL22e, but not FLAG-RpL22e-like-PA, is susceptible to phosphatase in vitro and smt3 (SUMO) knockdown in vivo. (A) Incubation of S2 cell and testis tissue extracts with calf-intestinal alkaline phosphatase in vitro significantly reduces immunodetection of the testis-specific 55γ RpL22e species (arrow). (B) Phosphatase treatment has no effect on the RpL22e-like-PA immunodetection pattern in testis. (C) In vivo knockdown of smt3 (via UAS-GAL4 binary system) was achieved by expressing a miR1 cassette targeting smt3 using a germline-specific GAL4 driver (bam-GAL4-VP16, UAS-Dicer2). Altered SUMOylated protein levels in the testis, determined by Western analysis, confirm smt3 knockdown. The testis RpL22e immunodetection pattern is significantly altered upon smt3 knockdown compared with control tissue. No change in RpL22e-like-PA accumulating levels is seen. In an attempt to produce a stronger smt3 knockdown, we used the early germline-specific GAL4 driver, nos-GAL4; however, smt3 depletion in this case results in complete loss of the germline (data not shown).

Figure 6. Modified RpL22e does not co-sediment with the translation machinery. (A) S2 cell extracts were separated in a 10–50% buffered sucrose gradient for polysome analysis to assess RpL22e co-sedimentation with ribosomal subunits, 80S monosomes and polysomes. In cells treated with the elongation inhibitor cycloheximide (black line), all modified RpL22e (43α, β and 55α) accumulates at the top of the gradient and only the unmodified 33 kD RpL22e co-sediments with the 60S large subunit, 80S monosomes and polysomes. Treatment with the chain terminator puromycin (causing a disruption of polysomes and accumulation of 80S monosomes; gray line) shifts the immunodetection pattern of unmodified RpL22e from polysomes to monosomes. Endogenous RpL23a was used was a positive control. (B) Deletion of fly-specific histone H1-like domain that harbors five putative SUMOylation motifs (ΔH1; residues 176−299) results in stable FLAG-tagged RpL22e protein (C-terminally tagged) in S2 cells. Full length (residues 1–299) is represented as WT. Tubulin was used as a loading control. (C) Polysome analysis of S2 cells expressing RpL22e-FLAG (full length) or RpL22eΔH1-FLAG shows an equal distribution pattern, as both were found to co-sediment with the translation machinery. The presence of FLAG-RpL22 in less dense “top” fractions may be attributable to high overexpression levels. Endogenous RpL23a was used as a positive control.

Figure 7. RpL22e family members are differently localized in the male germline. (A) The mitotic and post-mitotic germline are separated by a dashed line. Mitotic spermatogonia are proximal to the apical tip (asterisk), where germline stem cells are located and germline development begins. Post-mitotic primary spermatocytes (will further develop and enter meiosis I) are found distal to the apical tip. Immunofluorescence (used to localize RpL22e family members) reveals distinct localization patterns in the male germline. RpL22e-like-PA (green) is primarily cytoplasmic, with some subnuclear accumulation (presumably in the nucleolus) in mitotic germ cells. Strong cytoplasmic localization is seen meiotic spermatocytes. RpL22e (red) is primarily distributed in the nucleus. A punctate RpL22e localization is seen in the mitotic germline (thin arrow), but becomes more nucleoplasmic in post-mitotic cells (arrow with filled-in arrowhead). Co-localization of RpL22e-like-PA with RpL22e is only seen in the nucleus (presumably in the nucleolus) in mitotic spermatogonia (thin arrows, merge). (B) RpL22e (red) is also detected in somatic cyst cells (arrow) that surround spermatogonial cysts. RpL22e-like-PA (green) is a distinct germline marker. Although the anti-RpL22e-like Ab detects both spliced isoforms (-PA and -PB), staining intensity is consistent with the relative accumulation of RpL22e-like-PA (based on immunoblot analysis) compared with RpL22e-like-PB in the testis. RpL22e-like-PA is far more abundant (~10,000 × ) than RpL22e-like-PB in the testis.

Figure 8. RpL22e co-localizes with nucleolar markers in mitotic spermatogonia and early (post-mitotic) primary spermatocytes. GFP-tagged Nucleostemin1 (GFP-NS1) was expressed in vivo in the early germline using the germline-specific bam-GAL4-VP16 driver. Co-localization of GFP-NS1 (green) nucleolar foci with RpL22e (red) subnuclear immunostaining is evident in mitotic spermatogonia and early primary spermatocytes (arrows). Mitotic spermatogonia are proximal to the apical tip (asterisk), where germline stem cells are located and germline development is initiated. Post-mitotic primary spermatocytes are found distal to the apical tip.

Figure 9. RpL22e does not co-localize with the nucleolus in mature meiotic spermatocytes. (A) Schematic representation of male reproductive tract as shown in Kearse et al. (2011). Inset shows region of testis distal from the apical tip where populations of meiotic spermatocytes are represented. Immunohistochemistry of the RpL22e paralogues shows distinct punctate nuclear localization of RpL22e (red) in maturing primary spermatocytes, whereas RpL22e-like (green) remains primarily cytoplasmic (open arrows). The developmental gradient of germline maturation (from less mature to more mature) is represented by the long closed arrow. SV: seminal vesicle. (B) Phase contrast and immunohistochemistry of mature meiotic primary spermatoctyes shows juxtaposition, but not co-localization, of the phase dark nucleoli and RpL22e (red; arrows). DAPI staining (green) was used to confirm nuclear co-localization. (C) Co-staining of the nucleolar marker fibrillarin (green) and RpL22e (red) in mature meiotic primary spermatocytes confirms RpL22e is nucleoplasmic, not nucleolar (arrows). DAPI staining (blue) was used to confirm nuclear co-localization.

Figure 10. RpL22e localization in mature spermatocytes is sensitive to SUMO levels. In vivo knockdown of SUMO (smt3) was achieved by expressing a miR1 RNA cassette targeting smt3 (Fig. 5C). Nucleoplasmic RpL22e immunolocalization (red) in mature meiotic spermatocytes is generally widespread (thin arrows) as a result of the smt3 knockdown, as compared with control tissue (wildtype) where the RpL22e nucleoplasmic pattern is more punctate (arrows with filled-in arrowheads). DAPI staining is seen in blue.