Characterization of Aspergillus nidulans TRAPPs uncovers unprecedented similarities between fungi and metazoans and reveals the modular assembly of TRAPPII

Abstract

TRAnsport Protein Particle complexes (TRAPPs) are ubiquitous regulators of membrane traffic mediating nucleotide exchange on the Golgi regulatory GTPases RAB1 and RAB11. In S. cerevisiae and metazoans TRAPPs consist of two large oligomeric complexes: RAB11-activating TRAPPII and RAB1-activating TRAPPIII. These share a common core TRAPPI hetero-heptamer, absent in metazoans but detected in minor proportions in yeast, likely originating from in vitro-destabilized TRAPPII/III. Despite overall TRAPP conservation, the budding yeast genome has undergone extensive loss of genes, and lacks homologues of some metazoan TRAPP subunits. With nearly twice the total number of genes of S. cerevisiae, another ascomycete Aspergillus nidulans has also been used for studies on TRAPPs. We combined size-fractionation chromatography with single-step purification coupled to mass-spectrometry and negative-stain electron microscopy to establish the relative abundance, composition and architecture of Aspergillus TRAPPs, which consist of TRAPPII and TRAPPIII in a 2:1 proportion, plus a minor amount of TRAPPI. We show that Aspergillus TRAPPIII contains homologues of metazoan TRAPPC11, TRAPPC12 and TRAPPC13 subunits, absent in S. cerevisiae, and establish that these subunits are recruited to the complex by Tca17/TRAPPC2L, which itself binds to the 'Trs33 side' of the complex. Thus Aspergillus TRAPPs compositionally resemble mammalian TRAPPs to a greater extent than those in budding yeast. Exploiting the ability of constitutively-active (GEF-independent, due to accelerated GDP release) RAB1* and RAB11* alleles to rescue viability of null mutants lacking essential TRAPP subunits, we establish that the only essential role of TRAPPs is activating RAB1 and RAB11, and genetically classify each essential subunit according to their role(s) in TRAPPII (TRAPPII-specific subunits) or TRAPPII and TRAPPIII (core TRAPP subunits). Constitutively-active RAB mutant combinations allowed examination of TRAPP composition in mutants lacking essential subunits, which led to the discovery of a stable Trs120/Trs130/Trs65/Tca17 TRAPPII-specific subcomplex whose Trs20- and Trs33-dependent assembly onto core TRAPP generates TRAPPII.

Fig 1. Composition of TRAPPs by shotgun proteomics.

(A) Schemes depicting the subunit composition of TRAPP complexes and table showing the designation of each subunit in A. nidulans, metazoans and S. cerevisiae. To facilitate understanding we use protein designations from metazoan and yeast as appropriate throughout the text. (B) SDS-PAGE and silver staining analysis of TRAPPII purified with Trs120-S. (C) Abundance of TRAPP subunits in Trs120-S-purified material. Bars indicate spectral counts normalized to the number of Lys and Arg residues for each protein (i.e. the approximate number of tryptic peptides potentially detectable by MS/MS, multiplied by 100 and plotted as the mean +/- S.E.M. of N = 5 experiments). (D) SDS-PAGE of Bet5-S-purified TRAPPs. (E) Abundance of TRAPP subunits in Bet5-S-purified material (mean +/- S.E.M. of N = 5 experiments). (F) SDS-PAGE of Trs85-S-purified TRAPPs. (G) Abundance of TRAPP subunits in Trs85-S-purified material (mean +/- S.E.M. of N = 3 experiments).

Fig 2. LYRA analysis of essential genes.

(A) Heterokaryon rescue of lethal deletion alleles by constitutively active rab1* and rab11* alleles. Conidiospore suspensions were plated to give isolated colonies. For genotypes leading to lethal or severely debilitating phenotypes, conidiospores were obtained from heterokaryons. Wild-type colonies producing green or white conidiospores are shown on the right for comparison. (B) Color plot summarizing LYRA analysis. Dark blue represents wild-type growth whereas red indicates no growth; pink, very weak rescue; light blue, substantial rescue. (C) GDP release assays of wild-type and constitutively active D124E RAB1 (RAB1*). Time courses of mant-GDP exchange for GDP were used to calculate the observed k_off of GDP.

Fig 3. Gel filtration analysis of TRAPP complexes.

Lysates prepared from the wt or the indicated mutants expressing endogenously HA-tagged TRAPP proteins were run through a Superose 6 column whose fractions were analyzed by anti-HA western blotting. Levels of each protein in any given fraction are plotted as percentage of the total signal in the column. The resulting elution profiles were smoothened using the ‘Simple Spline Curve’ option of SigmaPlot’s Graph menu. Elution positions of protein standards (in kDa) are indicated on the top. The identity of the different peaks is shown. (A) Wild-type Trs23-HA3 extracts. (B) Wild-type Trs130-HA3 extracts; reference (black), wild-type Trs23-HA3. (C) Purified TRAPPII (Trs23-HA3); reference (black), wild-type Trs130-HA3 extracts. (D) trs65Δ Trs130-HA3 extracts; reference (black) wild-type Trs130-HA3 extracts. (E) Wild-type Trs85-HA3 extract; reference (black), wild-type Trs23-HA3. (F) Purified TRAPPIII (Trs23-HA3); reference (black), Trs85-HA3. (G) trs85Δ Trs23-HA3 extracts; reference (black), Trs23-HA3 in the wild-type. (H) trs120Δ rab11* Trs23-HA3 extracts; reference (black), Trs23-HA3 in the wild-type. (I) trs85Δ trs120Δ rab11* Trs23-HA3 extracts; reference (black), Trs130-HA3 in the wild-type. (J) Wild-type Trs20-HA3 extracts; reference (grey), wild-type Trs23-HA3. (K) Wild-type Tca17-HA3 extracts; reference (black), wild-type Trs130-HA3 extracts. (L) trs33Δ Trs23-HA3 extracts; reference (black), wild-type Trs23-HA3 extracts.

Fig 4. Negative staining 2D averages from purified TRAPP complexes.

Complexes eluted from S-agarose columns were concentrated, stained with uranyl acetate and analyzed to obtain 2D class averages. (A) TRAPPII purified with Trs120-S from wild-type, trs65Δ and tca17Δ cells. The percentage of particles in each class is indicated. (B) TRAPPIII purified with Trs85-S from wild-type cells. The position of the components, based on previous studies, is indicated [37]. (C) Left, TRAPPI purified with Bet5-S from trs85Δ trs120Δ rab11* cells. Right, an atomic model generated from pdb crystal structures 3CUE, 2J3W and 2J3T [3,6] was manually superimposed onto the TRAPPI 2D class average.

Fig 5. Composition of TRAPP complexes purified from trs120Δ, trs33Δ, trs31Δ and trs23Δ mutants analyzed by MS/MS.

Parallel purifications of TRAPP complexes from wild-type and mutant cells were carried out using Trs120-S or Trs65-S (for TRAPPII) and Bet5-S (for all TRAPPs). Proteins eluted from the S-agarose resin with S-peptide were analyzed by shotgun MS/MS. For each mutant condition, the normalized protein scores obtained as in Fig 1 (i.e. spectral counts/number of Arg + Lys residues, multiplied by 100) were plotted relative to the corresponding scores in the wild-type, which were set as 100%. Diagrams on the right schematically depict the composition of TRAPP complexes. (A) Bet5-S and Trs65-S in trs120Δ cells. (B) Bet5-S and Trs120-S in trs33Δ cells. (C) Bet5-S and Trs120-S in trs31Δ cells rescued with rab1* and rab11*. (D) trs23Δ cells rescued with rab1* and rab11*. Whether the incomplete core complexes shown in brackets in the trs31Δ and trs23Δ mutants actually exist cannot be determined from available data.

Fig 6. Composition of TRAPP complexes purified from trs20Δ and tca17Δ mutants: the TRAPPII-specific subcomplex.

MS/MS analysis of purified TRAPP complexes as in Fig 5 and detection of the TRAPPII-specific subcomplex by gel filtration chromatography. (A) Shotgun proteomics of Bet5-S and Trs120-S complexes purified from trs20Δ cells rescued by rab11*. (B) As in (A) from tca17Δ cells also rescued by rab11*. (C) Gel filtration profile of trs20Δ rab11* Trs130-HA3 cell lysates. (D) Gel filtration profile of tca17Δ rab11* Trs130-HA3 cell lysates. (E) Gel filtration profiles of trs20Δ rab11* Trs130-HA3 and tca17Δ rab11* Trs130-HA3 cell lysates combined in the same plot for comparison. Elution profiles were smoothened using the ‘Simple Spline Curve’ option of SigmaPlot’s Graph menu, as in Fig 3. The identity of the different peaks is indicated. Note that under these mutant conditions a substantial amount of material, likely representing aggregates, elutes with the excluded volume.