TRAFD1 (FLN29) Interacts with Plekhm1 and Regulates Osteoclast Acidification and Resorption - PubMed (original) (raw)

TRAFD1 (FLN29) Interacts with Plekhm1 and Regulates Osteoclast Acidification and Resorption

Hanna Witwicka et al. PLoS One. 2015.

Abstract

Plekhm1 is a large, multi-modular, adapter protein implicated in osteoclast vesicle trafficking and bone resorption. In patients, inactivating mutations cause osteopetrosis, and gain-of-function mutations cause osteopenia. Investigations of potential Plekhm1 interaction partners by mass spectrometry identified TRAFD1 (FLN29), a protein previously shown to suppress toll-like receptor signaling in monocytes/macrophages, thereby dampening inflammatory responses to innate immunity. We mapped the binding domains to the TRAFD1 zinc finger (aa 37-60), and to the region of Plekhm1 between its second pleckstrin homology domain and its C1 domain (aa 784-986). RANKL slightly increased TRAFD1 levels, particularly in primary osteoclasts, and the co-localization of TRAFD1 with Plekhm1 also increased with RANKL treatment. Stable knockdown of TRAFD1 in RAW 264.7 cells inhibited resorption activity proportionally to the degree of knockdown, and inhibited acidification. The lack of acidification occurred despite the presence of osteoclast acidification factors including carbonic anhydrase II, a3-V-ATPase, and the ClC7 chloride channel. Secretion of TRAP and cathepsin K were also markedly inhibited in knockdown cells. Truncated Plekhm1 in ia/ia osteopetrotic rat cells prevented vesicle localization of Plekhm1 and TRAFD1. We conclude that TRAFD1, in association with Plekhm1/Rab7-positive late endosomes-early lysosomes, has a previously unknown role in vesicle trafficking, acidification, and resorption in osteoclasts.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Binding of Plekhm1 with TRAFD1.

(A) Pull-down of full-length, N-terminally TAP-tagged (streptavidin binding protein:HA:calmodulin binding protein) Plekhm1 with TRAFD1-FLAG constructs overexpressed in HEK293 cells. Numbers of constructs in upper panel correspond to lanes in lower panel. Lysates were pulled down with streptavidin-Sepharose followed by western blotting with anti-FLAG (middle blot). The same membrane was probed with anti-HA antibody to monitor pull-down efficiency of Plekhm1 (upper blot). A fraction of the lysate was also probed with anti-FLAG to control for TRAFD1 protein expression (lower blot). Amino acids 37–60 (lane 2 and blue box), containing the zinc-finger, were required for binding to Plekhm1 (lane 2). (B) Reciprocal experiments were done with Plekhm1 constructs diagrammed in the upper panel, with numbers corresponding to lanes in the blots in the lower panel. IP of C-terminally FLAG-tagged full-length TRAFD1 was done with TAP-tagged Plekhm1 constructs. Lysates were immunoprecipitated with anti-FLAG-agarose, blotted, and probed with anti-HA to detect Plekhm1 (middle blot). The same membrane was stripped and probed with monoclonal anti-FLAG antibody to monitor immunoprecipitation efficiency (upper blot). A fraction of the lysate was also probed with anti-HA to control for Plekhm1 protein expression (lower blot). Amino acids 784–986 (of 1059 total), between the second PH domain and the C1 domain, were required for binding TRAFD1 (lane 4 and blue box). The experiments were performed at least 3 times and representative gels are shown. Numbers on left show positions of molecular weight markers. ZF = zinc finger; F = FLAG tag; T = TAP tag; R = RUN domain; * = PH domains; C = C1 domain.

Fig 2. Expression profile of TRAFD1 in mouse cells.

(A) Transcript level of Trafd1 in RAW264.7cells (light green bars) and mouse BMMC (dark green bars) was measured by Q-PCR in cells treated with LPS (100 ng/ml, 7 hours) or RANKL (10 ng/ml RAW264.7 cells, 20 ng/ml BMMC, 3 days). Untreated cells were used as a control. One-way ANOVA was carried out and values are mean + standard deviation (s.d.) of 3 independent experiments. *** _P<_0.0001 versus control group. (B) Protein expression of TRAFD1 in RAW264.7 and mouse bone marrow mononuclear cells (BM) treated as in A. Whole cell extracts (50 μg) were analyzed by 8% SDS-PAGE, blotted onto PVDF, and probed with anti-TRAFD1 antibody. Lamin B1 was used as a loading control. The experiments were performed at least 3 times and representative gels are shown.

Fig 3. TRAFD1 localization in mouse BMMC.

(A) Immunostaining for endogenous TRAFD1 and Plekhm1 was performed on mouse monocytes and on mouse BMMCs treated with RANKL cultured on glass (osteoclasts), or on Osteo Assay plates (resorbing osteoclasts). Confocal microscopy images were obtained using anti-Plekhm1 and anti-TRAFD1 (colors as indicated), and representative images are shown. DAPI staining was used to visualize nuclei. Scale bars = 10 μm. Insets show enlarged regions outlined in white. Arrowheads indicate examples of vesicles where co-localization gives yellow signal. Pearson’s correlation coefficient was used to estimate the co-localization of TRAFD1 with Plekhm1, where 1 represents perfect co-localization and 0 is random co-localization. One-way ANOVA was carried out and values are mean + s.d. of n = 3 independent experiments analyzing at least 4 images/condition. *** P<0.0001 versus monocytes. (B) Immunostaining for endogenous Plekhm1 and Rab7 and co-localization analysis was performed as in A. Scale bars = 10μm. Values are mean + s.d. of n = 3 independent experiments analyzing at least 4 images/condition. No significant change in co-localization was observed ± RANKL or resorption. M = monocytes; OCs = osteoclasts; res OCs = resorbing osteoclasts.

Fig 4. TRAFD1 knockdown inhibits resorbing activity of osteoclasts.

(A) Stable knockdown of TRAFD1 was achieved by transduction with lentivirus expressing shRNA (or scrambled control) in RAW264.7 cells. Clones were selected with different degrees of knockdown in the absence of RANKL, as shown (clones 1.7, 2.3; and 2.6). Q-PCR experiments were performed at least 3 times and representative graph shows means+ s.d. of duplicates. *** P<0.0001 versus control group. (B) Resorption assay was performed with the knockdown clones in A. Cells were cultured on 24-well Osteo Assay (HA) plates for 10 days in the presence of RANKL (10 ng/ml). Cells were removed and the plate was scanned at high resolution. Representative wells are shown on lower panel (black = resorbed area). The percentage of resorbed area over total area of well is indicated on upper panel. Analysis of resorbed area was measured by Image J software. One-way ANOVA was carried out and values are mean +s.d. of n = 3 independent experiments analyzing 3 wells/condition. *** P<0.0001 versus control group.

Fig 5. shTRAFD1 cells differentiate in vitro.

(A) Cells stably expressing TRAFD1 shRNAs were cultured on 96-well plates in the presence of RANKL (10 ng/ml), fixed, and stained for TRAP on the days indicated. Representative micrographs are shown. Scale bar = 100 μm. (B) Rhodamine phalloidin and DAPI staining of osteoclast-like cells cultured on glass coverslips. shTRAFD1 and control RAW264.7 cells were cultured in the presence of RANKL (10 ng/ml) on 24-well plates with glass coverslips on the bottom. When cells differentiated (3 days for controls, 4 days for shTRAFD1), they were fixed and stained. Representative fluorescence micrographs are shown. Scale bar = 10 μm.

Fig 6. Decreased level of TRAFD1 has minor effects on expression of osteoclast resorption factors.

(A) Q-PCR analysis is shown for the indicated genes in differentiating shTRAFD1 cells (4 days in RANKL) vs. control cells (3 days in RANKL). The expression of each gene in shTRAFD1 cells is normalized to the expression of control cells on day 3. Student’s _t_-test was carried out and values are mean +s.d. of triplicate determinations in 3 independent experiments. A representative graph is shown. * P<0.05; *** P<0.0001 versus control. (B) Whole cell protein extracts from differentiated control cells (3 days in RANKL) and differentiated shTRAFD1 cells (4 days in RANKL) were subjected to immunoblotting using the antibodies shown. The experiments were performed at least 3 times and representative gels are shown.

Fig 7. Knockdown of TRAFD1 inhibits acidification of osteoclasts.

Acidification assay was performed with shTRAFD1 and control RAW264.7 cells. Cells were cultured on 24-well Osteo Assay plates for 5 days in the presence of RANKL (10 ng/ml) and stained with acridine orange. Representative confocal microscopy images are shown. White rectangles locate areas shown at higher magnification in insets, below. As a control for blocked acidification, some control cells were treated with bafilomycin A (200 nM; control + baf A). Scale bar = 50 μm.

Fig 8. TRAFD1 is required for TRAP and cathepsin K secretion.

(A) Differentiated RAW 264.7 cells and shTRAFD1 cells were grown for 5 days on Osteo Assay plates in the presence of RANKL. Conditioned media were collected 2 and 5 days post differentiation and TRAP secretion was measured biochemically. A representative graph is shown. One-way ANOVA was carried out and values are mean +s.d., n = 8 per time point. *** P<0.0001 vs. controls. (B) TRAFD1 knockdown and control RAW264.7 cells were cultured on Osteo Assay plates in the presence of RANKL. Culture supernatant (sup) was collected 24 hours post differentiation. 40 μl of conditioned media were subjected to western blot analysis for the presence of cathepsin K. Total cell lysates (TCL) from each well served as total protein controls. The same gel was re-probed with lamin B1 antibody. The experiments were performed 3 times and representative gels are shown.

Fig 9. TRAFD1 in osteoclast centrifugal fractionation is Plekhm1 dependent.

Osteoclasts from WT and ia/ia (Plekhm1 truncation) osteopetrotic rats were osmotically lysed and fractionated by differential centrifugation. All fractions were subjected to western blot. 50 μg of protein from each fraction were loaded per lane. TCL: total cell lysate; NF: nuclear fraction and cell debris; LMF: light mitochondrial and lysosomal fraction; VF: vesicle fraction; CF: cytosolic fraction. The membrane was probed with anti-Plekhm1, anti-TRAFD1, anti-Rab7 (late endosome/early lysosome marker), and anti-lamin B1 (nuclear envelope marker). Truncated Plekhm1 is seen in ia/ia rats. The experiments were performed at least twice and representative gels are shown.

References

1. Witwicka H, Hwang S-Y, Odgren PR (2014) The Structure of Bone Reference Module in Biomedical Sciences: Elsevier.
1. Li YP, Chen W, Liang Y, Li E, Stashenko P (1999) Atp6i-deficient mice exhibit severe osteopetrosis due to loss of osteoclast-mediated extracellular acidification. Nat Genet 23: 447–451. - PubMed
1. Rajapurohitam V, Chalhoub N, Benachenhou N, Neff L, Baron R, Vacher J (2001) The mouse osteopetrotic grey-lethal mutation induces a defect in osteoclast maturation/function. Bone 28: 513–523. - PubMed
1. Sobacchi C, Schulz A, Coxon FP, Villa A, Helfrich MH (2013) Osteopetrosis: genetics, treatment and new insights into osteoclast function. Nat Rev Endocrinol 9: 522–536. 10.1038/nrendo.2013.137 - DOI - PubMed
1. Boyle WJ, Simonet WS, Lacey DL (2003) Osteoclast differentiation and activation. Nature 423: 337–342. - PubMed

TRAFD1 (FLN29) Interacts with Plekhm1 and Regulates Osteoclast Acidification and Resorption - PubMed (original) (raw)