The subcellular localization of TRPP2 modulates its function

Abstract

TRPP2, also known as polycystin-2, is a calcium permeable nonselective cation channel that is mutated in autosomal dominant polycystic kidney disease but has also been implicated in the regulation of cardiac development, renal tubular differentiation, and left-to-right (L-R) axis determination. For obtaining further insight into how TRPP2 exerts tissue-specific functions, this study took advantage of PACS-dependent trafficking of TRPP2 in zebrafish larvae. PACS proteins recognize an acidic cluster within the carboxy-terminal domain of TRPP2 that undergoes phosphorylation and mediate retrieval of TRPP2 to the Golgi and endoplasmic reticulum (ER). The interaction of human TRPP2 with PACS proteins can be inhibited by a Ser812Ala mutation (TRPP2(S812A)), thereby allowing TRPP2 to reach other subcellular compartments, and enhanced by a Ser812Asp mutation (TRPP2(S812D)), thereby trapping TRPP2 in the ER. It was found that the TRPP2(S812A) mutant rescued cyst formation of TRPP2-deficient zebrafish larvae to the same degree as wild-type TRPP2, whereas the TRPP2(S812D) mutant was significantly more effective in normalizing the distorted body axis of TRPP2-deficient fish. Surprisingly, the TRPP2(S812D) mutant rescued the abnormalities of L-R asymmetry more effectively than either wild-type or TRPP2(S812A), suggesting that the ER localization of TRPP2 plays an important role in the development of normal L-R asymmetry. Taken together, these findings support the hypothesis that TRPP2 assumes distinct subcellular localizations to exert tissue-specific functions.

Figure 1.

Sequence comparison of the acidic cluster of TRPP2 and the furin binding regions of PACS-1 and -2. (A) The acidic clusters (boxed) are conserved between human and zebrafish TRPP2/polycystin-2 but are lacking in C. elegans TRPP2. (B) Human PACS-1 furin-binding region (FBR; NM_018026) and zebrafish PACS-1 FBR (EF531598). (C) Human PACS-2 FBR (NM_015197) and zebrafish PACS-2 FBR (EF531599). The human and zebrafish PACS-1 and -2 FBR are highly conserved. ClustalW and boxshade were used for the alignment.

Figure 2.

The FBR of zPACS-1/2 interacts with zebrafish TRPP2 and human TRPP2. The interaction is enhanced for TRPP2^S812D but reduced for TRPP2^S812A. Zebrafish TRPP2 is retained in the ER. (A) HEK 293T cells were transiently transfected with the flag-tagged cytoplasmic domain of zebrafish TRPP2 (F9.zTRPP2.cyt) and with V5-tagged FBR of zPACS-1 and -2 (V5.zPACS-1 or -2.FBR). F9.zTRPP2.cyt co-precipitated with V5.zPACS-1/2.FBR but not with V5.GFP. (B) Transient transfection of HEK 293T cells with V5.zPACS-1.FBR and flag-tagged cytoplasmic domain of human TRPP2 (F9.hTRPP2.cyt) WT or mutated serine 812 (WT, S812A, S812D). V5.zPACS-1.FBR bound and co-precipitated the C-terminus of human TRPP2 WT and TRPP2^S812D but TRPP2^S812A only weakly. (C) In a similar experiment, V5.zPACS-2.FBR interacted with the C-terminus of human TRPP2 WT and TRPP2^S812D but less strongly with TRPP2^S812A. Similar results were obtained by pull-down experiments (Supplemental Figure 2). (D) Overexpression of F9.zTRPP2 and BAP31.EGFP in HeLa cells shows an overlapping distribution of both proteins in the ER. (E) Overexpression of F9.zTRPP2 in zebrafish PAC2 cells (kept at 28°C) shows a corresponding overlap with calnexin in the ER.

Figure 3.

Overexpression of zPACS-1 and zPACS-2 by injection of capped RNA induced cyst formation in zebrafish. Zebrafish PACS-1 or PACS-2 capped RNA or the combination of both was injected into zebrafish eggs at one-cell stage; at 55 hpf, the injected larvae were subdivided into dysmorphic groups (d1 through d3) according to the gross morphologic changes as shown in A. (A) The phenotype induced by PACS overexpression was reminiscent of the zpkd2 knockdown phenotype (Figure 4B): Body axis curvature (double arrow), hydrocephalus (arrowhead), pericardial edema (double arrowhead), and pronephric cyst formation (arrow, confirmed by histology) are marked. (B) The images on the left side show a WT control larva; the pronephros is hardly discernible with transmitted-light microscopy (top left). On histologic cross-sections, the glomerulus appears normal (white arrow, bottom left). In injected larvae, the pronephric tubules were distended to cysts (white arrows, top right), which are more clearly visible on histologic cross-section (*, bottom right); the glomerulus was stretched (black arrow, bottom right). (C) The appearance rate of the overall dysmorphic features was dosage dependent (data not shown). The embryos were injected with 50 (zPACS-2) or 180 (zPACS-1) pg capped RNA, respectively; the percentages of the dysmorphic groups (d1 through d3) and lethality are shown in the bar graph. (D) The rate of cyst formation was higher with overexpression of zPACS-2, which interacts more strongly with zTRPP2 (Figure 2A) compared with zPACS-1. The combination of zPACS-1 and zPACS-2 showed a small additive effect on the rate of dysmorphic changes of the injected larvae (C) but not on cyst formation (D). Retention of zTRPP2 at the ER (zPACS-2) and the Golgi (zPACS-1) seems to be necessary for induction of cyst development. The number n in D equals n in C without the dead embryos.

Figure 4.

zpkd2 splice morpholinos induce pronephric cysts in zebrafish. Overexpression of WT human TRPP2, TRPP2^S812A, and TRPP2^S812D in zebrafish infrequently leads to pronephric cyst formation. (A) WT larva (55 hpf) with normal histology of the pronephric glomerulus (white arrow) and tubules (white arrowhead). (B) The effect of polycystin-2 depletion induced by injection of 13 ng zpkd2 splice morpholino (pkd2MO) consists of dorsal body curvature, hydrocephalus (arrowhead), and pronephric cyst formation (arrow; 55 hpf). Histologic section revealed dilated pronephric tubules (*) and a stretched glomerulus in the midline (black arrow). (C) Human TRPP2^WT, TRPP2^S812A (SA), and TRPP2^S812D (SD) were overexpressed by capped RNA injection (200 pg per embryo at one-cell stage). The injected larvae showed similar phenotypes and were classified into three groups according to the degree of their dysmorphic changes (d1 through d3). Body axis curvature (double arrow), hydrocephalus (arrowhead), and pericardial edema (double arrowhead) are marked. The histology of the pronephros was preserved in most larvae, but approximately 1% of d2/d3 classified zebrafish larvae developed pronephric cysts. (D) Quantitative distribution of the dysmorphic classes (d1 through d3) in the differently injected groups: The overall number of WT-looking larvae is equal, although hTRPP2^S812A seems to cause higher lethality than hTRPP2 and hTRPP2^S812D. Lower amounts of capped RNA (100 pg; see Figure 5) had only minor or no detectable effects.

Figure 5.

Rescue of the zpkd2 morphant phenotype by co-injection of capped mRNA of WT human TRPP2 and mutant hTRPP2^S812A and hTRPP2^S812D. All larvae were examined at 55 hpf. (A) The zpkd2 morphant larvae had a dorsally bent body axis, hydrocephalus (arrowhead), and pronephric cysts (arrow), as described in Figure 4. Co-injection of pkd2MO with 100 pg of hTRPP2 WT capped RNA inhibited the formation of pronephric cysts, hydrocephalus, and body axis curvature (judged by trunk-tail angle of <90°, because most of the morphant larvae showed an angle of considerably more than 90°). Co-injection of pkd2MO with hTRPP2^S812A (SA) mRNA showed a similar effect as WT mRNA. Co-injection of pkd2MO with hTRPP2^S812D (SD) capped RNA could not inhibit cyst formation as efficiently as WT hTRPP2 or hTRPP2^S812A capped RNA (arrow) but had a stronger rescue effect on the body curvature. (B) Graphical presentation of the rescue experiments by co-injection of pkd2MO and 100 pg of hTRPP2^{(WT, S812A, S812D)} capped RNA compared with injection of pkd2MO alone with regard to cyst formation and body axis curvature (judged by trunk-tail angle of <90°): The rescue with hTRPP2^WT was statistically significant in both aspects of the evaluation (* in C); a rescue to the same extent was seen with hTRPP2^S812A, whereas hTRPP2^S812D had a significant effect only on body axis curvature. (C) Net effect of rescue relative to the percentage of phenotype: There was no difference in the rescue of cyst formation between hTRPP2^WT (31.2%) and hTRPP2^S812A (34.3%) mRNA, but the degree of rescue was less with hTRPP2^S812D mRNA (10.5%). With regard to the rescue of the body axis, there was also no significant difference between hTRPP2^WT and hTRPP2^S812A mRNA, but the effect of hTRPP2^S812D RNA (43.4%) that is held back in the ER/Golgi by the interaction with PACS molecules was significantly better. Data of seven independent experiments were pooled. Compared with injection of morpholino alone, statistically significant differences obtained by applying the χ² test are marked with *.

Figure 6.

Differential effects of the rescue experiments on L-R patterning. The larvae were fixed and processed for antisense in situ experiments at stage 55 hpf. (A) Expression pattern of the used in situ markers. The expression of the heart muscle marker myl7 (myosin, light polypeptide 7; A, top left) highlights a normal position of the ventricle (v) and atrium (a) in an uninjected control larva; in a ventral view, the ventricle loops to the right and the atrium loops to the left. Many pkd2MO-injected embryos showed a inversed situs (A, top right). The gastrointestinal marker gene forkhead-2 (fkd2) highlights the liver (li) being positioned to the left of the midline and the pancreas (p) to the right side in a dorsal view of control larva (A, bottom left); for better visualization of the pancreas, the pancreatic islet marker preproinsulin (ins) was used in addition to fkd2. In many pkd2 morphants, the situs of these two organs was inversed (A, bottom right). (B) Disturbance of L-R axis formation by injection of pkd2MO can lead to midline situs or bilateral formation of an internal organ. For example, a heart in the midline (B, top), a liver in the midline, and a pancreas with a right-sided position (B, middle) or bilaterally located pancreata with a inversely positioned liver (B, bottom). (C) Organ situs frequency distribution: Situs solitus (normal organ position), situs inversus (inversed position), and midline or bilateral organ situs. Position of the heart, liver, and pancreas was judged separately. Compared with WT larvae, pkd2MO-injected larvae had a situs inversus of heart, liver, and pancreas in approximately 40% and midline/bilateral position of approximately 10%. (D) Percentage of abnormal situs by averaging the numbers of the three organs being looked at per group. The one-tailed t-test was used to compare the rescue groups versus the morpholino only group. (E) Percentage indicates relative rescue compared with pkd2MO alone. Among the groups co-injected with pkd2MO and capped RNA, the order of magnitude of rescue effect was mutant hTRPP2^S812A (SA RNA) < WT hTRPP2 (WT RNA) < hTRPP2^S812D (SD RNA). The co-injection with SD RNA had the most pronounced effect, the situs abnormalities decreased by approximately 30%, and this rescue effect was significant compared with pkd2 morphant larvae and rescue with WT RNA. The data are from three independent experiments.