Summary- paper 56: ER calcium depletion as a key driver for impaired ER-to-mitochondria calcium transfer and mitochondrial dysfunction in Wolfram syndrome

Mailis Liiv, Annika Vaarmann, Dzhamilja Safiulina, Vinay Choubey, Ruby Gupta, Malle Kuum, Lucia Janickova, Zuzana Hodurova, Michal Cagalinec, Akbar Zeb, Miriam A. Hickey, Yi-Long Huang, Nana Gogichaishvili, Merle Mandel, Mario Plaas, Eero Vasar, Jens Loncke, Tim Vervliet, Ting-Fen Tsai, Geert Bultynck, Vladimir Veksler & Allen Kaasik

Nature Communications, 2024

Questions/gaps addressed:

• Wolfram syndrome is a rare genetic disease caused by mutations in the wolframin gene (WFS1)-prevalent, or in the CDGSH iron-sulfur-containing domain 2 gene (CISD2)-rarer. Both genes encode transmembrane endoplasmic reticulum (ER) proteins.

• loss of WFS1 reduced the number of mitochondrial contact sites with the ER (MAMs), decreased Ca2+ uptake by mitochondria, and decreased mitochondrial respiration

Major hypotheses:

Key methods:

Major takeaways:

Summary- paper 55: Dynamic movement of the Golgi unit and its glycosylation enzyme zones

Akihiro Harada, Masataka Kunii, Kazuo Kurokawa, Takuya Sumi, Satoshi Kanda, Yu Zhang, Satomi Nadanaka, Koichiro M. Hirosawa, Kazuaki Tokunaga, Takuro Tojima, Manabu Taniguchi, Kenta Moriwaki, Shin-ichiro Yoshimura, Miki Yamamoto-Hino, Satoshi Goto, Toyomasa Katagiri, Satoshi Kume, Mitsuko Hayashi-Nishino, Miyako Nakano, Eiji Miyoshi, Kenichi G. N. Suzuki, Hiroshi Kitagawa & Akihiko Nakano

Nature Communications, 2024

Questions/gaps addressed:

• The distribution of glycosylation enzymes in the Golgi apparatus has been well examined for cis-trans distribution but where do they localise within a single cisterna?

Key methods:

• CRISPR-Cas9 knock-in for endogenous tagging
• super-resolution confocal live imaging microscopy SCLIM, immunoelectron microscopy.
• Enzymes studied: Early stage:
  • N-glycosylation enzyme, MGAT1 (Alpha-1,3-Mannosyl-Glycoprotein 2-Beta-N-Acetylglucosaminyltransferase),
  • O-glycosylation enzyme, GALNT6 (Polypeptide N-Acetylgalactosaminyltransferase 6)
  • GAG synthesising enzyme, XYLT2 (Xylosyltransferase 2) Late stage:
  • β4GalT1 (beta4-galactosyl transferase 1), an enzyme synthesising polylactosamine on both N- and O-glycans
  • NDST1 (N-Deacetylase And N-Sulfotransferase 1), which catalyses the transfer of sulphate to the nitrogen of glucosamine in a GAG molecule, heparan sulphate
• Golgi discs as the Golgi ‘units’: dispersed the Golgi by nocodazole, which dissociates the large Golgi apparatus into smaller Golgi discs, or ‘mini-stacks’ surrounded by giantin or other golgins (golgin-84 and TMF1).
• RUSH construct of syndecan2 (SDC2), a substrate protein for GAG modification to look at kinetic signature of colocalization with early and late glycoslation enzymes.

Major takeaways:

• Mainly observational. Glysocylation enzymes loaclize to specific ‘zones’. Treatment with or not with nocodazole does not change the observed zones. The glycosylation substrates enter the glycosylation zones according to the order of the glycosylation reaction.

Summary- paper 54: TFasting-induced RNF152 resensitizes gallbladder cancer cells to gemcitabine by inhibiting mTORC1-mediated glycolysis

Ying Tao,Zijun Gong,Sheng Shen,Yaqi Ding,Rui Zan,Bohao Zheng,Wentao Sun,Chaolin Ma, Mengxuan Shu,Xiao Lu, Han Liu,Xiaoling Ni,Houbao Liu,and Tao Suo

iScience, 2024

Questions/gaps addressed:

• Gallbladder cancer has poor prognosis, often detected after metastasis. Fasting leads to reduced proliferation of cancer cells presumably by inhibiting AKT/mTOR signaling to prevent the Warburg shift and enhance cancer cell sensitivity to chemotherapy.

• RNF152 was found in a genome-wide CRISPR-Cas9screen to identify genes involved in gemcitabine resistance in Gallbladder cancer. RNF152 is downregulated in several biliary tract cancers. What is the relationship between RNF152, mTORC1, p18/LAMTOR1, and gemcitabine resistance?

Major hypotheses:

• Fasting upregulates the expression of RNF152 which targets p18/LAMTOR1 to inhibit the activity of mTORC1, reducing the activity of p-4E-BP1 and pS6K1, and thereby increase the sensitivity of gallbladder cancer cells to gemcitabine.

Key methods:

• Human Gallbladder cancer cell lines (GBC-SD and SGC-996).

• Short-term starvation (STS): fasting mimic medium (glucose-free DMEM supplemented with 0.5 g/L glucose and 1% FBS) for 48 h. Normal control media: DMEM with 10% fetal bovine serum.

Major takeaways:

• When RNF152 is inhibited, levels of p18/LAMTOR1 increase. Addition of proteasome inhibitor, MG132 (10 mM), but not lysosome inhibitor, bafilomycin A1 (BafA, 100 nM, a V-ATPase inhibitor), stabilized p18, suggesting p18 is degraded by the proteasome. Lower levels of p18 also correlate with, lower mTORC1 activity, lower glycolysis, and reduced cell growth.

Summary- paper 53: The ER calcium channel Csg2 integrates sphingolipid metabolism with autophagy

Shiyan Liu, Mutian Chen, Yichang Wang, Yuqing Lei, Ting Huang, Yabin Zhang, Sin Man Lam, Huihui Li, Shiqian Qi, Jia Geng, Kefeng Lu

Nature communications, 2023

Questions/gaps addressed:

• Calcium conc in the cytosol is low (0.1 µM), extracellular calcium at 1 mM, intracellular calcium reservoirs (endosome 2 µM, lysosome at 0.5 mM, ER at 0.5 mM, Golgi at 0.1 mM, mitochondria at 0.1 µM and secretory vesicle at 0.1 mM). cytosolic calcium conc is rapidly modulated by ion channels and pumps. What are the consequences of abnormal calcium accumulation in reservoir organelles caused by dysfunctional calcium channels?

• Cytosolic calcium binds and activates calmodulins that transduce signals into appropriate downstream outputs through calcineurin phosphatase and kinases, regulating pathways such as autophagy. What is the mechanism by which calcium regulates autophagy?

• Sphingolipids are ubiquitous components of eukaryotic membranes and function as signaling bioactive lipids regulating cell growth, endocytosis, protein trafficking etc. What mechansims regulate sphingolipid homeostasis?

Major hypotheses:

• Calcium and sphingolipid homeostasis are interlinked and regulate autophagy.

Key methods:

• Screened calcium channel mutants in yeast (csg2Δ, pmr1Δ, spf1Δ, pmc1Δ, vcx1Δ, yvc1Δ, mid1Δ, rch1Δ, ecm7Δ, cch1Δ) for viability after amino acid starvation.
  • Csg2 was the most defective for growth after 72h in SD-N (mild phenotype for Ecm7, Mid1).
  • Defective for GFP-Atg8, GFP-50Q model reporter processing after 16h in SD-N, 16h in SD-Glu. RFP-Ape1 and Pho8Δ60 failed to be delivered to vacuole lumen for 3h in SD-N.
  • No defect in Ape1 maturation in nutrient rich conditions (although Atg1 is blocked).
  • Pmr1 deletion led to upregulation of autophagy (GFP-Atg8 processing observed within 6h in SD-N). Do transport directions of calcium (into the ER by Pmr1 and out of the ER by Csg2) cause opposite effects on autophagy?

• To monitor the stage of autophagy that is blocked, looked at GFP-Atg8 localization: diffuse cytosolic distribution of Atg8 indicating the failure of autophagosome fusion with vacuoles. Deletion of the SNARE subunit Vam3 or GTPase Ypt7 lead to accumulation of GFP-Atg8 puncta. No defect in general vacuolar delivery- monitored by CPY-mCh, GFP-Sna3.

• Csg2 mainly predicted to contain 10 TMDs and a potential calcium binding EF-hand motif. Csg2-GFP, Csg2-HA, and HA-Csg2 complement function. Csg2-GFP localizes to the ER.

• Purified Csg2 protein, reconstituted into a planar lipid bilayer and measured Ca channel activity using a single-channel conductance assay. Did not observe conductance for Mg, Zn, Mn, Fe ions.

• A genetic fluorescent probe of ER calcium, Stt3-jGCaMP7c to measure ER Calcium levels and using a kit using ER calcium probe Mag-Fluo-AM. Higher ER calcium levels in Csg2-deleted mutant. jGCaMP7c is a sensitive sensor with low background fluorescence.

• Functionally important regions: Deletion of any TMD killed function. Deletion of the EF-hand motif had not effect on function or on ER localization.

• Treatment of csg2∆ cells with 100 mM CaCl2 (up to 800 mM), did not rescue autophagy defect.
  • Treatment with EDTA (5 mM), or EGTA (20 mM), restored autophagy. Increase in ER calcium rather than a decrease responsible for autophagy block in csg2∆.
  • Deleting Csg2 and Pmr1 together (but not vacuolar calcium channels Pmc1, Vcx1, or Yvc1) also restored autophagy.
  • Deletion of the cytosolic calmodulin-calcineurin-Crz1 signaling components Cna1, Cnb1, and Crz1 did not rescue autophagy
• Comprehensive lipidomics analysis in WT vs csg2∆ under starvation.
  • phytoCers and PHS (immediate precursors of complex sphingolipids) exhibited significant accumulation in the mutant.
  • Complex sphingolipid mannosylinositol phosphorylceramide (MIPC) was reduced in csg2∆.
  • Sphingolipid synthesis pathway genes tested. Tsc10, Lip1, Aur1. No effect of Tsc10, but Lip1 and Aur1 mutants blocked autophagy. Double mutant csg2∆ tsc10-ts restored autophagy. No effect of combining Csg2 and Lip1/Aur1.
  • Steady state stability of Aur1-GFP/ GFP-Aur1 reduced in csg2∆ during starvation. But the levels of Kei1 (in complex with Aur1 for IPS synthesis) not affected.
    • Treatment with EGTA/EDTA rescued Aur1 levels.
    • Aur1 also stable in csg2∆ pmr1∆ double mutant (so dependent on ER calcium levels).
    • Aur1 stable when treated with proteasome inhibitor MG132 but not by the vacuole inhibitor NH4Cl, so degraded by the proteosome.
  • Tor1 activity is higher in csg2∆ during starvation (monitored by Sch9 phosphorylation) also inhibiting autophagy.

Major takeaways:

• Deletion of Csg2, the ER-resident calcium channel, causes accumulation of calcium in the ER, which disturbs sphingolipid metabolism by destabilizing Aur1 (how?), leading to increases in PHS that block autophagy (how?). Mechanism of PHS regulated autophagy inhibition unclear.

Summary- paper 52: Golgi-IP, a tool for multimodal analysis of Golgi molecular content

Rotimi Fasimoye, Wentao Dong, Raja S. Nirujogi, Eshaan S. Rawat, Miharu Iguchi, Kwamina Nyame, Toan K. Phung, Enrico Bagnoli, Alan R. Prescott, Dario R. Alessi, Monther Abu-Remaileh

PNAS, 2023

Questions/gaps addressed:

Major hypotheses:

Key methods:

• Bait used: uncharacterized Golgi membrane protein TMEM115-3xHA (yeast homolog YOL107W).

Major takeaways:

Summary- paper 50: The wtf4 meiotic driver utilizes controlled protein aggregation to generate selective cell death

Nicole L Nuckolls, Anthony C Mok, Jeffrey J Lange, Kexi Yi, Tejbir S Kandola, Andrew M Hunn, Scott McCroskey, Julia L Snyder, María Angélica Bravo Núñez, Sarah E Zanders

eLife, 2020

Questions/gaps addressed:

• wtf4 gene is a meiotic driver in Schizosaccharomyces pombe and S kombucha that uses a poison-antidote mechanism to selectively kill meiotic products (spores) that do not inherit wtf4. Spores expressing just the toxin die, while those expressing both the toxin and antidote survive. How do these drivers work and selectively destroy spores?

• Wtf4 poison and antidote are encoded from overlapping coding sequences, alternate splicing leads to an additional start codon and the antidote contains ~45 additional N-terminal amino acids, the C terminal 293 aa are the same. The two proteins are predicted to contain 6 TMDs.

Key methods:

• Express S. kombucha wtf4-poison and antidote proteins in S. pombe and in vegetative S. cerevisiae. Inducible over-expression of the poison coding sequence (using the Z3EV β-estradiol inducible system) kills the cells/ spores. Co-expression of the antidote coding sequence restores viability.

• Fluorescence microscopy, FRET, Distributed Amphifluoric FRET (DAmFRET), half punctum-Fluorescence Recovery After Photobleaching (half-FRAP): show poison localizes in punctate structures, antidote in peri-vacuolar puncta. When co-expressed, both are in the peri-vacuolar structures or inside the vacuole. Poison and the antidote are close enough to physically interact, and likely self-assemble/interact with themselves. When photobleached, do not see recovery up to 3 min, suggesting the assemblies are not very dynamic.

• Increasing the repeats of 7 amino acid sequence (IGNAFRG) that is found in varying copy numbers in many members of the wtf gene family, in the wtf4 poison or antidote alone, disrupted the interaction, but together restored the neutralization ability.

• Transmission electron microscopy (TEM): Wtf4 antidote localizes to vesicles adjacent to the vacuole. By fluorescence microscopy that is adjacent to Rnq1 and Atg8 puncta- the authors suggest these to be IPOD structures. They also report accumulation of lipid droplets in cells expressing the toxin-antidote. Also observe mitochondriala hyper-fission.

• Unbiased SGA screen to identify mutants that don’t grow when the toxin-antidote are expressed- to identify what is the machinery required to neutralize these. Identify several mitochondrial proteins, CWI pathway, ESCRTs, CORVET, Vps1, vacuolar and endosome trafficking machinery. In the absence of many of these, even the antidote by itself is toxic.

Major takeaways:

• Wtf4 toxin and antidote are membrane proteins, that differ in the N terminal 45 aa. Toxin by itself localizes to dispersed puncta, antidote by itself is peri-vacuolar. Together they are both peri-vacuolar. Why is the toxin toxic and not the antidote? Is it something about the difference in the N term or that the antidote is more effectively cleared due to its peri-vacuolar localization? Why is the localization different- is the N terminus the driver for the localization?

• Mitochondria, and the ESCRT pathway are two main ways of clearing these toxin-antidote intermediates.

Summary- paper 43: Novel Mechanism for Tubular Injury in Nephropathic Cystinosis

Swastika Sur, Maggie Kerwin, Silvia Pineda, Poonam Sansanwal, Tara K. Sigdel, Marina Sirota, Minnie M. Sarwal

Biorxiv, 2022

Questions/gaps addressed:

• Cystinosis is a rare autosomal recessive LSD, caused by mutations in the CTNS gene. The kidney is the first organ affected functionally. The most severe form is nephropathic/infantile cystinosis, resulting in end-stage renal disease (ESRD) by 10 years of age. What is the mechanism of defects in renual tubular cells, are these due to functions besides cystine transport?

Major hypothesis:

• Defective lysosomal clearance may be a pivotal cause of the altered function of renal proximal tubular epithelial cells (RPTECs).

Key methods:

• whole-genome microarray transcriptional expression profiles in RPTECs (isolated from patient urine samples) and fibroblasts (isolated from the same patient skin biopsies) from patients with cystinosis and healthy age- and gender-matched controls.

Major takeaways:

• vATPase subunits levels are affected in CTNS KO

Summary- paper 42: The P4-ATPase Drs2 interacts with and stabilizes the multisubunit tethering complex TRAPPIII in yeast

Irene Pazos, Marta Puig-Tintó, Laura Betancur, Jorge Cordero, Nereida Jiménez-Menéndez, Marc Abella, Altair C Hernández, Ana G Duran, Emi Adachi-Fernández, Carla Belmonte-Mateos, Susana Sabido-Bozo, Sébastien Tosi, Akiko Nezu, Baldomero Oliva, Julien Colombelli, Todd R Graham, Tamotsu Yoshimori, Manuel Muñiz, Maho Hamasaki, Oriol Gallego

EMBO Reports, 2023

Questions/gaps addressed:

• Eight Multisubunit tethering complexes (MTCs) have been described in yeast, all conserved in humans: COG, Dsl1, CORVET, HOPS, GARP, exocyst, TRAPPII and TRAPPIII.

• What are the broad interaction partners of multiple MTCs?

Key methods:

• Mining published genetic interaction datasets for multiple MTCs. Selected 426 proteins for subsequent protein–protein interaction test by PICT assay (why use genetic interaction datasets for this shortlist?).

• PICT (Protein interactions from Imaging of Complexes after Translocation) assay based on the rapamycin-induced heterodimerization of the FK506-binding protein (FKBP) and the FKBP-rapamycin binding (FRB) domain in a tor1-1 bkg. The anchor is at the spindle pole body (Tub4-RFP-FKBP). This assay cannot distinguish between direct and indirect interactions. PICT screen identified seven candidates, one of which is Drs2 which binds both TRPAII and III.

• Tested interaction between Drs2 and other P4-ATPases and all the MTCs again. Multiple MTCs bound Drs2.

• TRAPIII is required for Atg9 trafficking to the PAS, and hence needed for Cvt pathway. Drs2 is essential for cell growth below 21°C. Tested if Drs2 is required for Cvt (measured by Ape1 precessing by WBs)? Yes, but only at low temp, while Trs85 (TRAPIII) is required at all temps.

• Electron microscopy (EM) to look at Cvt flux. Don’t detect any Cvt body-like structures in the vacuoles of pep4∆ drs2∆.

• Correlative Light and Electron Microscopy (CLEM) to look at Ape1-GFP aggregates in drs2∆. No Ape1-GFP aggregates in the vacuole of drs2∆ cells.

• Drs2 is a lipid flippase is involved in multiple endosome and Golgi transport steps. Blocking transport from early endosomes to TGN (rcy1∆), blocking AP-1 pathway from TGN to early endosomes and the exit of the high-density class of vesicles from the TGN (apl4∆), blocking ALP pathway from TGN to the vacuole (apl5∆), or blocking CPY pathway from TGN to late endosomes (gga1 gga2) do not disrupt Ape1 processing. So, none of these roles of Drs2 likely impair Ape1 processing.

• Drs2 activity mutants: PS flipping defective (drs2-GA), ATPase-dead (drs2-D560N) show no defect in Ape1 processing. No defect in cho1∆ (required for synthesis of PS) also had no defect- so PS is not required for the Cvt pathway.

• C terminal tail mutants of Drs2: LSSLVMK motif-Gea2 interacting motif (GIM), another conserved motif (modulates autoinhibitory conformation), two NPFXD motifs endocytosis signals: none of these affect Ape1 processing.

• N terminal tail mutants of Drs2: found a conserved stretch of 15 amino acids I(S/R)TTK motif (residues 198–212 in Drs2) in all the P4- ATPases bind MTCs (Dnf1, Dnf2, and Drs2) but absent in Neo1 and Dnf3. The stretch defines a cavity of 18.8 Å in the solved Drs2 structure. drs2-5A mutant lacking the motif grows at low temp, but is defective in Ape1 processing.

• Competition assay by overexpressing the N tail of Drs2 results in some inhibition of Ape1 processing even in the presence of WT Drs2.

• Cross-linking-mass spectrometry (XL-MS) immunoprecipitation to look at Drs2 interactors (wt vs 5A mutant)- disuccinimidyl sulfoxide (DSSO), a cross-linker of 10.1 Å in length. drs2-5A-GFP destabilized interaction with Trs85 (TRAPIII), Trs65, Trs120, Trs130 (specific subunits of TRAPPII) and Vps53 (GARP).

• Live cell microscopy to track Atg9 (fast and slow moving puncta). In drs2∆ and drs2-5A cells, the low mobility population increases.

Major takeaways:

• Drs2 and other P4 ATPases can function as regulatory partners of MTCs. MTCs bind the N-terminal cavity of P4-ATPases that nest the somewhat conserved I(S/R)TTK motif in the N tails, and this mechanism may be conserved.

Summary- paper 34: Lysosome transporter purification and reconstitution identifies Ypq1 pH-gated lysine transport and regulation

Felichi Mae Arines, Aleksander Wielenga, Olive E. Burata, Francisco Narro Garcia, Randy B. Stockbridge, Ming Li

Biorxiv, 2023

Questions/gaps addressed:

• Eukaryotic membrane proteins are usually expressed at low levels, don’t express well in bacteria due to loss of post-translational modifications, and are often unstable in detergent based purifications. Development of more methods to express, purify, and reconstitute lysosomal transporters into liposomes for biochemical characterization.

Key methods:

• Comparison of protein overexpression: YP+Gal > YNB+Gal. Also pTEF> pGPD> pADH. pTEF and YP +Gal are comparable.

• 4 copies of Protease cleavable construct: pTEF1::Ypq1-HRV 3C protease site-GFP-8xHis cloned in integrative pRS305 and pRS306 backbones in both mating types. Generated diploids to maximize overexpression. Protein stability improved if using GFPS65T instead of yEGFP (yEGFP→GFPS65T: G65T, A72S, Q80R).

• Cell lysis and purification: 4L YPD cultures, grown to high OD. Spheroplasted with zymolyase. Dounce homogenized. Solubilized proteins in 2% DDM. Affinity chromatography 8xHis with TALON cobalt resin. Cleavage with HRV 3C protease, followed by purification by size exclusion chromatography.

• Purified Ypq1 in +Lysine conditions, when incubated in buffer- lysine, leads to instability and protein aggregation. Progressive reduction in lysine conc: Ypq1 stable in 10 mM and 1 mM, but not in 0.1 mM lysine. Reported lysine pools inside the vacuole (∼5-10 mM) and the cytoplasm (∼0.5-0.8 mM).

• Reconstitution into proteoliposomes and transport: reconstituted Ypq1 into yeast polar lipid extract (Avanti Polar Lipids) in the presence of 10 mM lysine in neutral internal buffer (pH 7.2). Exchanged the proteoliposome into an external buffer containing 2 μM 14C-lysine and ATP.

• Maximal uptake when intra proteoliposome pH is 6.2 and outside pH is 8.2 with a lysine gradient.

Major takeaways:

• Massive overexpression does not result in constitutive degradation of Ypq1 in vivo.

• Lysine and Arginine transport possible. Both can compete in the proteolipsosome based transport assay. But in vivo, removal of arginine does not lead to Ypq1 degradation.

• PQ loop mutant (P229S,Q230R) resistant to degradation upon lysine withdrawal. Also stable in vitro +/- lysine. PQ mutant also defective in protein transport. PQ motif acts as a molecular hinge that allows PQ-loop transporters to change conformations during transport.

Summary- paper 32: The K/HDEL receptor does not recycle but instead acts as a Golgi-gatekeeper

Jonas C. Alvim, Robert M. Bolt, Jing An, Yasuko Kamisugi, Andrew Cuming, Fernanda A. L. Silva-Alvim, Juan O. Concha, Luis L. P. daSilva, Meiyi Hu, Dominique Hirsz & Jurgen Denecke

Nature Communications, 2023

Questions/gaps addressed:

• Golgi membrane-spanning sorting receptors mediate the sorting of cargo but also need to be sorted themselves.

• KDEL receptor/ Erd2 is believed to be recycled between ER and Golgi, but those studies were based on C-terminal tagged Erd2. Does Erd2 really cycle between Golgi and ER?

Key methods:

• Localization of AtErd2-YFP is mostly ER with some Golgi puncta, but the construct is not functional. YFP-TM-AtErd2 (YFP followed by an extra N terminal TMD) is mostly at the Golgi and is functional. Same localization dynamics when expressed the tagged At or HsErd2 constructs in human cells.

• α-amylase-HDEL (Amy-HDEL) secretion assay: All Erd2 orthologs with greater than 50% similarity reduced secretion of Amy-HDEL. Same phenotype for untagged or YFP-TM-Erd2 construct but not the Erd2-YFP construct.

• Mutation of a “cluster of lysines” near the ERD2 C-terminus proposed to form the retrieval motif for COPI coat for Golgi to ER transport, has no effect on function. Mutation of conserved leucines at position −3 and −5 from the C-terminus (note yeast Erd2 does not have L at -5) killed Erd2 function.

• Disruption of LxLPA C terminal motif led to Erd2 redistribution to the ER that increased when more ligand was co-expressed. Why does LLGG mutant bind COPI to recycle back to ER instead of travelling further in the Golgi?

• Tested other small tags at the C terminus of At or Hs Erd2: FLAG, c-myc, HA. FLAG and c-myc were non-functional and alos showed ER localization. HA tag was functional and localized to the Golgi.

• High overexpression of WT but not ∆C5 Erd2 mutant, similar to BFA treatment, led to mixing of ER-Golgi and caused Arf1-RFP to become cytosolic, and prevented secretion of both Amy and Amy-HDEL. How high expression of Erd2 leads to complete collapse of the ER-Golgi system is unclear?

• How does the low expression of Erd2 as compared to cargo proteins explain its function? If Erd2 is not packaged in the COPI retrograde carrier, what provides specificity to the cargo being packaged in the COPI vesicles?

Major takeaways:

• YFP or RFP fused to the ERD2 C-terminus inactivate the receptor. Using YFP followed by an extra TMD (from ERP1 which is a PQ family protein but contains 8 TMDs) at the N-terminus seems to be the best tag of the combinations tested.

• Exposure of a native C-terminus di-leucine motif (LXLPA) in Erd2 is important for its Golgi localization. How conserved is the dileucine motif? What does it bind to?

• ERD2 cycles between ligand-bound and ligand-free configurations within the Golgi, but does not cycle between Golgi and the ER. How does Erd2 define the specifity of packaging of its cargos without being packaged itself in the COPI vesicle?

Summary- paper 31: Distinct role of TGN-resident clathrin adaptors for Rab5 activation in the TGN-endosome trafficking pathway

Makoto Nagano, Kaito Aoshima, Hiroki Shimamura, Daria Elisabeth Siekhaus, Junko Y. Toshima, Jiro Toshima

Biorxiv, 2023

Questions/gaps addressed:

• Two types of endosomal compartment in yeast: an early endosome-like compartment containing Rab5 (Vps21p) and a late endosome-like compartment containing Rab7. What is the identity of the Vps21p-residing compartment and how is the compartment generated?

• GGAs and AP-1 form two distinct clathrin-coated vesicles, the GGA-enriched vesicle, including Ent3 and a minor population of Ent5, and the AP-1-enriched vesicle, which includes most of Ent5 and resides at the late TGN.

• Activation of yeast Vps21 (Rab5) is partially suppressed by deletion of the epsin-like proteins, Ent3p/5p. Deletion of GGAs or AP-1 complex, or even both, has little effect on Vps21p activity.

• What are the functional relationships between clathrin adaptors in the Rab5 activation in yeast?

• Arf1 GTPase and Frq1p, recruit Pik1 to the TGN, which generated PI4P, which recruite GGA first, then AP-1 and Ent5. Ypt31p/32p (Rab11) also regulate the localization of Ent3/5 at the TGN. So Rab11 and Pik1 both regulators?

Key methods:

• Synthetic genetic interactions test between ent3Δ or ent5Δ mutation with deletion of the GGA1/2 or APL4 (AP-1γ subunit) and looked at Vps21/Rab5 localization. Decrease of GFP-Vps21p fluorescence intensity and an increase in the number of Vps21p compartments when combined ent5Δ with gga1Δ/2Δ. So, GGA-dependent recruitment of Ent3p is important for formation of the Vps21p compartment. AP-1, as well as Ent3p/5p, play a role in formation of the Vps21p compartment.

• Method to quantitatively assess the Rab5 activity using Nano luciferase fused to either a nanobody that recognizes GFP or to a fragment of human Rabenosyn-5 that binds specifically to the GTP-bound form of Vps21p. Vps21p tagged with GFP-ALFA peptide. Pull from the cell extracts with GST-fused ALFA nanobody, monitoring luciferase activity allows looking at the ratio of active Vps21 to total Vps21. Significantly reduced active Vps21 in ent3Δ ent5Δ apl4Δ vs wt. AP-1 functions redundantly with Ent3p/5p in Vps21p activation.

• GFP-Vps21 accumulates in clusters near the vacuole in vps9∆, and in ent3Δ ent5Δ apl4Δ. Hse1p (endosomal marker) but not Sec7p (TGN marker) localizes at the Vps21p cluster adjacently to the vacuole. Endosomal characteristics of this cluster? Suppressed by expression of constitutively active Q66L mutant of Vps21p.

• Role of GGAs, PI(4)P, and yeast Rab11 homologues Ypt31p/32p in the localization of Ent3p/5p and AP-1 at the TGN. GGA1 and GGA2 mutants decreased Ent3 at the TGN, no effect on Ent5 and Apl2. Temperature-sensitive pik1-1(D1055G) mutant, affected all three Ent3, Ent5 and Apl2 (strongest). Rab11 ypt31ts mutant has stronger effect on Ent3, Ent5, weaker on Apl2.

• Transport of Vps9p from the TGN to the endosome is necessary for Vps21p activation. Shift in punctuate localization of GFP-Vps9p from endosomes to the TGN in the ent3Δ ent5Δ apl4Δ mutant, leading to less active Vps21 at the endosome, and less E-dots. Disruption of Ent3, Ent5 , and Ap-1 progressively suppresses transport of Vps9p from the TGN to the Vps21p compartment. (see model in Fig 7).

Major takeaways:

• GGA vesicles are formed approximately 10 s prior to the AP-1 vesicles, and GGA vesicles mediate the transport of cargos from the TGN to the vacuole. AP-1 vesicles mediate intra-Golgi recycling.

• All the three clathrin adaptors, Ent3p/5p, AP-1 and GGAs, contribute to Rab5/Vps21 activation.

• Ent3p/5p and AP-1 are required for loading Vps9p on the vesicles that transport Vps9p to the Vps21p compartment, while GGAs, together with Pik1p and Rab11s, have a role in the recruitment of Ent3p/5p and AP-1 to the TGN.

• PI(4)P (Pik1p), Rab11 (Ypt31p/32p), and GGAs are involved in the recruitment of Ent3p to the TGN, function cooperatively recruit Ent3p and contribute to the activation of Vps21p

• Ent5p is able to complement the function of Ent3p in Rab5 activation. Does Ent5p have dual roles at the TGN: loading of Vps9p onto transport vesicles at the GGA-enriched early TGN, and recycling at the AP-1-enriched late TGN?

• Recently reported, Tlg2p-residing compartment of the TGN, is distinct from the late TGN where Sec7p resides, is the first destination for endocytic traffic, from which the endocytic cargo is transported to the Vps21p compartment.

Summary- paper 27: Mechanism of signal-anchor triage during early steps of membrane protein insertion

Haoxi Wu, Ramanujan S. Hegde

Molecular Cell, 2023

Questions/gaps addressed:

• Key step in the biogenesis of a membrane protein is the insertion of its TMDs into the lipid bilayer in the correct topology. How is this coordinated? Key players involved: Sec61 translocon, EMC insertase, SRP and SRP receptor.

• Factors thought to be important:
  • for about half of the membrane proteins, the first TMD serves as a signal sequence (signal anchor) with a preceding N-tail.
  • how is it decided whether the N tail is in the lumen or cytosol? has been atrributed to flanking positive charge in the cytosol, signal anchor length and hydrophobilicity
• After co-translational targeting of an signal anchor to the ER, those intended for an N(exo) topology are inserted by EMC whereas N(cyt) are inserted by Sec61 translocon. This paper interrogates this further.

Key methods:

• Model substrate used: N(exo) signal anchor from trace-amine-associated receptor 5 (TAAR5), a G-protein-coupled receptor inserted by EMC.

• In vitro translation of 35S-labeled TAAR5-SA in the presence of semi-permeabilized cells. Addition of a glycoslation site at the N terminus allowed monitoring glycosylation as a proxy for translocation. strongly EMC dependent ( blocked in EMC6 KO) and mostly Sec61 independent (no effect of Apratoxin A (Sec61 inhibitor)).

• Translation stalled at a defined site using rare codons, the ribosome-nascent-chain complexes (RNCs) purified by sucrose-gradient fractionation, followed by incubation with semi-permeabilized cells for 10 min. Translocation of stalled at Signal anchor +50 and 70 aa strongly EMC dependent and mostly Sec61 independent. Signal anchor +30aa was Sec61 dependent and EMC independent.

• Used cysteine-reactive chemical crosslinking with bismaleimidohexane (BMH): produced sucrose-gradient-purified 35S-labeled TAAR5-SA+70 RNCs containing a single cysteine in the N-tail and crosslinking with BMH.

Major takeaways:

• A two-step sequential triage by EMC and Sec61 mediates the insertion and topogenesis of signal anchors.

• Membrane protein topology is not determined only after RNC delivery to Sec61. Found an additional step between SRP-mediated targeting and ribosome docking at Sec61. EMC transiently samples signal anchors at an earlier step and inserts a subset of them in the N(exo) topology. Those signal anchors that are skipped by EMC arrive at Sec61, which preferentially favors SA insertion in the N(cyt) topology.

Summary- paper 26: Yeast Svf1 binds ceramides and contributes to sphingolipid metabolism at the ER cis-Golgi interface

Sergej Limar, Carolin Körner, Fernando Martínez-Montañés, Viktoriya G. Stancheva, Verena N. Wolf, Stefan Walter, Elizabeth A. Miller, Christer S. Ejsing, Vanesa Viviana Galassi, Florian Fröhlich

Journal of Cell Biology, 2023

Questions/gaps addressed:

• Ceramides are synthesized in the ER and receive their head-groups in the Golgi apparatus, yielding complex sphingolipids (such as sphingomyelin in mammalian cells or inositol-containing ceramides in yeast). Transport of ceramides between the ER and the Golgi is performed by the ceramide transport protein (CERT) in mammalian cells. CERT is not conserved in yeast, how is ceramide transport accomplished in yeast?

• Several candidates for ceramide transport in yeast have been proposed: non-vesicular: Nvj2, Tricalbins, OSBP homologs Osh2, Osh3, and Osh4, and vesicular trafficking. Exact mechanism and contributions unclear.

Major hypotheses:

• Since Nvj2-mediated ceramide transport occurs only under conditions of high ER stress, hypothesized that other transporter proteins are likely involved.

Key methods:

• Analyzed high-throughput chemical genetics (Fitdb) dataset to identify genetic similarity to Nvj2. Found Svf1 shows high correlation with Nvj2, Osh1, Osh3.

• Tetrad analysis and negative genetic interaction: Between svf1∆ and osh2Δ osh3Δ osh4Δ; and nvj2∆ and osh2Δosh3Δosh4Δ. svf1∆ nvj2∆ sensitive to myriocin (0.5 µM).

• Fluorescence microscopy and localization of C-term Svf1-GFP (colocalized with cis-Golgi marker Mnn9-mKate, and cytoplasm). N-term tagged GFP-Svf1 is not functional.

• HeliQuest analysis to identify N-terminal membrane targeting motif. Found a potential amphipathic helix (with one face made of hydrophobic residues and the other one containing polar residues).

• Mutating the hydrophobic valine (position 12) to negatively charged aspartic acid, largely destroying the hydrophobic moment, led to loss of cis-Golgi puncta. The mutant also appears to be less stable, and unable to complement function in svf1∆ sur2∆ mutant.

• Tested requirement of N-terminal acetylation for membrane insertion (similar to that reported for Grh1) by mutating the second aa (L2E), or using a mutant lacking MAK3, the catalytic subunit of NatC acetyl transferase. Used mass spec of the N-terminus of Svf1 from Svf1-GFP pull down to monitor N-terminally acetylation.

• Minimal fusion constructs of the first 18 aa of Svf1 tagged with GFP localize to the ER (no cis-Golgi). Addition of amphipathic helix of Grh1 does not recue Golgi localization of function.

• SILAC MS to identify interactors in Svf1-GFP immunoprecipitates. Found several ER and early-Golgi membrane proteins.

• Tested the “hug-and-kiss” model for Svf1 at cis-Golgi and ER exit sites contact at the same time- ∼40% of Svf1 dots co-localized with both Mnn9-Kate (cis-Golgi) and Sec31-Halo (ER-exit sites).

• Tested whether COP-II is required for Svf1 targeting to the Golgi, Svf1-GFP in temperature-sensitive alleles of Sec23 (Sar1 GTPase-activating protein) or Sec17 (α-SNAP co-chaperone). Saw a reduction with cis-Golgi, but inconclusive.

• Measured the levels of yeast sphingolipids using LC-MS/MS. Also measured content of ceramides on in vitro budded COP-II vesicles. Saw decrease in complex sphingolipids, observed additive effects when combined with sec12 ts (COPII) mutant.

• Deletion of Svf1 in mutants lacking diacylglycerol acyltransferases (dga1Δ lro1Δ are1Δ are2Δ)- needed for detoxification of ceramides, led to high sensitivity to IPC synthase inhibitor Auroebasidin A (AbA) (added to further increase ceramide toxicity).

• AlphaFold structural prediction of Svf1 predicts two lipocalin domains made of anti-parallel β-sheets. Used AutoDock Vina molecular docking to predict ceramide binding sites. Two histidine residues located in a small α-helical stretch that lies over the potential ceramide binding pocket. Also mutated to alanines (H273A/H274A). The mutant lost ceramide binding and Golgi localization, and shows genetic interaction with sur2∆.

• Tested binding to ceramides, immunoprecipitated overexpressed (from pGal promoter) FLAG tagged Svf1, Svf1V12D, and Svf1G7A/G8A without detergents. Extracted lipids from the immunoprecipitates, determined the amount of co-purified ceramide 44:0;4 and, as a control, 34:1 phosphatidylcholine (PC 34:1) using targeted lipidomics. Observed enrichment of 44:0;4 ceramide for all WT and mutants- so amphipathic helix not important for ceramide binding.

Major takeaways:

• Identified the yeast protein Svf1 is a novel ceramide-binding protein. Svf1 deletion leads to accumulation of ceramides and decrease in complex sphingolipids. Show Svf1 binds ceramides, but how is the binding and release cycle regulated?

• Also possible that Svf1 does not directly transport ceramides but acts as a sensor for ceramides which then regulate downstream Aur1 activity in the Golgi apparatus.

Summary- paper 23: Proximity-labeling chemoproteomics defines the subcellular cysteinome and inflammation-responsive mitochondrial redoxome

Tianyang Yan, Ashley R. Julio, Miranda Villanueva, Anthony E. Jones, Andréa B. Ball, Lisa M. Boatner, Alexandra C. Turmon, Stephanie L. Yen, Heta S. Desai, Ajit S. Divakaruni, Keriann M. Backus

Biorxiv, 2023

Questions/gaps addressed:

• Cysteines in proteins are sensors of redox changes and many approches look at global snapshots of cystine oxidation states in specific conditions. However these approaches do not provide organelle specific redox responses.

• ROS generation in mitochondria is especially high, and function of mitochondrial proteins is especially affected in pro-inflammatory programs in macrophages. What is the extent of mitochondrial cysteine oxidation in these states?

Key methods:

• TurboID proximity based biotinylation with cysteine redox state analysis to enable in situ subcellular cysteine fractionation and quantitative measurement of cysteine oxidation state: Local Cysteine Capture (Cys-LoC).
  • Transiently overexpression of TurboID fusion proteins with localization sequences targeted to cytosol, endoplasmic reticulum, Golgi, mitochondrial, and nucleus.
  • TurboID proximal proteins are first biotinylated in situ.
  • Lysis and cysteine capping with the highly reactive iodoacetamide alkyne (IAA).
  • Biotinylated proteins enriched on streptavidin resin and subjected to sequence specific proteolysis to release all IAA-tagged peptides derived from the TurboID modified proteins.
  • Peptide-level click conjugation to biotin-azide followed by a second enrichment on neutravidin resin to capture of biotinylated cysteine peptides derived from TurboID modified proteins, followed by MS/MS.

Major takeaways:

• See organelle variability in the efficiency of the Cys-LoC pipeline working- works well for nucleus and cytosol, less effective for ER/ Golgi/ mito. Increasing labeling time (to 1h) and adding exogenous biotin to 500 µM improves efficiency somewhat.

• The biotinylation is promiscuous (observed by streptavidin-rhodamine), even though Mito-TurboID-eGFP reporter was mostly mitochondrial (observed by co-localization with mitotracker-tetramethylrhodamine ethyl ester (TMRE)).

Summary- paper 22: Cryo-EM structures of human SPCA1a reveal the mechanism of Ca2+/Mn2+ transport into the Golgi apparatus

Zhenghao Chen, Satoshi Watanabe, Hironori Hashida, Michio Inoue, Yasukazu Daigaku, Masahide Kikkawa, Kenji Inaba

Science Advances, 2023

Questions/gaps addressed:

• The mechanism of Ca2+ homeostasis in the Golgi apparatus which also acts as an intracellular calcium storage compartment, is less well understood than that in the ER. SERCA (Sarco(endo)plasmic Ca2+–ATPase) localizes to the ER and cis-Golgi, SPCA1 (Ca2+/Mn2+-ATPase) is present in the trans-Golgi and TGN for Ca2+ uptake, generating a large Ca2+ gradient concentration between the ER (~500 μM) and Golgi (~100 μM). How are SERCA and SPCA1 activities regulated?

• Loss-of-function mutations in the ATP2C1 (gene encoding SPCA1) causes an inherited autosomal dominant disease, Hailey-Hailey disease, a skin disorder characterized by persistent blisters. Homozygous KO of the ATP2C1 gene in mice is embryonic lethal due to failed neural tube closure. What is the structural basis of the defects in keratocytes and neuroepithelial cells in SPCA1 LOF?

• SPCA1 transports both Mn2+ and Ca2+ into the Golgi lumen, with a higher affinity for the former. Mn2+ is a crucial cofactor for many enzymes, and too much can also be toxic. Structural basis for difference in Mn2+ and Ca2+ transport by SPCA1?

Key methods:

• PA-tag (GVAMPGAEDDVV) and a tobacco etch virus (TEV) cleavage site at the N terminus of SPCA1 isoform 1a, expressed in HEK293T using the PiggyBac Cumate Switch Inducible system, induced with cumate (150 μg/ml) and 50 nM phorbol 12-myristate 13-acetate (PMA) for overexpression, solubilization in 1% DDM, and purification with anti–PA-tag Antibody Beads.

• SPCA1a is ubiquitylated at Lys8 (A-domain) and Lys496 (N-domain). The purified protein was deubiquitinated by treatment with Usp2cc enzyme to prepare a homogeneously deubiquitinated SPCA1a sample for cryo-EM.

• EnzCheck phosphate assay to monitor in vitro ATPase (and by proxy transport activity) of purified SPCA1. ATPase activity observed with multiple divalent metal cations- Co2+ and Ni2+, as well as with Ca2+ and Mn2+.

• To increase density of the A-domain of SPCA1, tested human cofilin-1 (hCFL-1) as a binder protein since it is known to bind the P-domain of SPCA1a but observed no interaction between purified hCFL-1 and SPCA1a in vitro assay using size exclusion chromatography. Instead focussed on stabilization by binding to a nanobody

• Flow cytometry screen for a nanobody that binds specifically and tightly to SPCA1a using a previously published yeast surface-displayed nanobody library (McMohan et al 2018, NSMB), found Nb14 to bind the N-domain of SPCA1a. Converted Nb14 to a megabody by fusing a scaffold protein to Nb14 while retaining the binding affinity for SPCA1a.

• To compare TM helix interactions between SERCA1a, SERCA2b, and SPCA1a, counted the number of H-bonds formed between different TM helices.

Major takeaways:

• The physiological significance of SPCA1a ubiquitylation is unclear.

• cryo-EM structures of human SPCA1a in two intermediate states at 3.1- to 3.3-Å resolution:
  • the E1-ATP state using β,γ-methyleneadenosine 5′-triphosphate (AMPPCP), a nonhydrolyzable analog of ATP, in the presence of Ca2+ or Mn2+, and
  • the E2P state using BeF3−, a phosphate group mimic, in the absence of these metal ions.

• SPCA1a shares the typical architecture of P-type ATPases, with 10 TM helices (TM1 to TM10) and three cytosolic domains, including the actuator (A), nucleotide-binding (N), and phosphorylation (P) domains. The TM domain of SPCA1a consists of three TM helix clusters (TM1–2, TM3–4, and TM5–10), similar to SERCA.

• The overall arrangement of the cytosolic and TM domains in SPCA1a resembles that in SERCA and that the location and amino acid residue geometry of the Ca2+/Mn2+-binding site in SPCA1a are nearly superimposable with those of the second Ca2+-binding site (site II) in SERCA2b.

• While the cytosolic domain arrangement is similar between SPCA1a and SERCA, the mode of interactions at the cytosolic domain interfaces is different between these two Ca2+ pumps. The Ca2+/Mn2+-binding site is primarily constituted by TM4 and TM6. Mn2+ occupies the same binding site in SPCA1a as Ca2+.

• Difference between SPCA1a and SERCA2b in TM2. TM2 is directly linked to the A-domain, the observed twist around the cytosolic end of TM2 in SPCA1a suggests that TM2 is one of the key elements that modulate the metal ion transport by this Ca2+/Mn2+ ATPase. TM2 twisting seems likely to be an essential event of several metal transporting P-type ATPases (SPCA1a, SERCA1, ATP13A, lipid P4-ATPase). Larger conformational and positional flexibility of TM2 and TM6 in SPCA1a compared with SERCA may help to explain the wider metal selectivity of SPCA1a.

• Appears that smaller numbers of H-bonds are formed between TM6 and its neighboring TM helices in SPCA1a (5 interdomain H-bonds with neighboring TMs), compared to those in SERCA1a and SERCA2b (10 and 11 H-bonds).

• The mode of ATP binding in SPCA1a is highly similar to that in SERCA2b.

• Hailey-Hailey disease associated mutations: most mutations occur around the metal-binding pocket and in the P-domain, suggesting that these mutations affect the metal transport activity, efficiency of ATP hydrolysis, and/or phosphorylation of Asp350 of SPCA1a.

• The Q747A gain-of-function mutation enhanced the SPCA1-mediated uptake of Mn2+, while the corresponding Q783A mutation in yeast PMR1 blocked Mn2+ transport. Why?

Summary- paper 21: Protein complexes in cells by AI-assisted structural proteomics

Francis J O’Reilly, Andrea Graziadei, Christian Forbrig, Rica Bremenkamp, Kristine Charles, Swantje Lenz, Christoph Elfmann, Lutz Fischer, Jörg Stülke, Juri Rappsilber

Molecular Systems Biology, 2023

Questions/gaps addressed:

• Although AI tools such as Alphafold multimer are invaluable, predicting the interaction interfaces of all possible combinations of protein pairs is prohibitively expensive and computationally impractical. Proteins also form complexes involving much larger numbers of subunits. How can we leverage the power of AI tools to benefit experimental data analysis?

• Existing validation tools for PPIs (eg, two-hybrid, affinity purification mass spectrometry (AP-MS) and co-fractionation MS studies (CoFrac-MS)) provide little topological or structural information, often even leaving open if an interaction is direct or indirect.

Major hypotheses:

• Combining complementary techniques including in-cell crosslinking to discover high-confidence direct protein interactions without genetic modification and to accurately predict and validate corresponding structural models can provide richer PPI information with higher accuracy.

Key methods:

• Generated a whole-cell interaction network using crosslinking and co-fractionation mass spectrometry in B. subtilis with the membrane-permeable crosslinker DSSO. Cells were treated with crosslinker, lysed, the proteins fractionated and trypsin digested, and the resulting peptides separated by cation exchange and size exclusion chromatography prior to mass spectrometry.

• Generated a comprehensive PPI candidate list (predictions from crosslinking MS, CoFrac-MS, and SubtiWiki) for system-wide structure modeling with AlphaFold-Multimer (version 2.1), added known PPIs that lack structural information to our experimentally identified PPIs.

• Models were assessed for overall predicted TM score (pTM) and interface predicted TM score (ipTM). pTM reports on the accuracy of prediction within each protein chain and ipTM on the accuracy of the complex. A TM-score of 0.5 is broadly indicative of a correct fold/domain prediction, while scores above 0.8 correspond to models with matching topology and backbone path. An ipTM > 0.85 has proven reliable in other analyses when compared to known interface TM score and the DockQ docking quality score.

• Employed a noise model in which 300 B. subtilis proteins from the datasets were predicted as pairs with random E. coli proteins. The ipTM distribution of the resulting decoy PPIs was compared with 10 subsamples of our AlphaFold-Multimer predictions. No decoy PPIs reported an ipTM > 0.85.

Major takeaways:

• Describe genetic-free approaches for protein–protein interaction screening capable of producing large numbers of novel protein–protein interactions along with their topologies by fixing interactions in cells. The experimental approaches yielded 44 high-quality PPI models (ipTM > 0.85).

• Crosslinking MS providing the highest “hit rate” for structural modeling of novel PPIs. 12% of crosslinking MS PPIs lead to models with ipTM > 0.85, 4% of CoFrac-MS, and 11% of the SubtiWiki dataset.

• Found a strong correlation between ipTM and restraint satisfaction of heteromeric crosslinks, despite the fact that crosslinking information was not used in AlphaFold model prediction. Crosslink violation is especially low with ipTM > 0.85, indicating that high-confidence models agree with the residue–residue distances observed in situ.

• A total of 80 of 153 high-quality dimers include proteins with transmembrane domains, and membrane proteins are present in half of the predicted trimer structures.

Summary- paper 18: Predominant Golgi Residency of the Plant K/HDEL Receptor Is Essential for Its Function in Mediating ER Retention

Fernanda A.L. Silva-Alvim, Jing An, Jonas C. Alvim, Ombretta Foresti, Alexandra Grippa, Alexandra Pelgrom, Thomas L. Adams, Chris Hawes, and Jurgen Denecke

The Plant Cell, 2018

Questions/gaps addressed:

• How do the cascade of interactions between KDEL ligands, ERD2, G-proteins, and protein kinase A, explain the transport of K/HDEL proteins back to the ER?

Key methods:

• Fusion protein Amy-HDEL (Barley α-amylase (Amy)) as cargo. Cotransfection with increasing amounts of ERD2a reduced the partial secretion of Amy-HDEL.

• Used a hybrid/ fusion ERD2 construct (NbERD2ab)- N term and 4 TMs of Erd2a + 4 TM and C term of Erd2b. Fusion effective in retention of Amy-HDEL. The two forms are interchangeable.

• Fluorescent reporters: ST-YFP or ST-YFP-HDEL. Agrobacterium tumefaciens-mediated transient expression in infiltrated tobacco leaf epidermis cells followed by confocal laser scanning microscopy. ST-YFP predominantly at the Golgi, ST-YFP-HDEL retained in the ER. Overexpression of Amy-HDEL caused a partial redistribution of ST-YFP-HDEL to the Golgi.

• N- and C-terminal YFP tagged Erd2: Erd2-YFP at eR+ Golgi. YFP-Erd2 predominantly ER. Both defective in Amy-HDEL retention. Tagging kills Erd2 function. Addition of N-terminal signal peptide and a short decapeptide harboring an N-linked glycosylation site, secYFP-Erd2, is only at the Golgi but is non-functional.

• Addition of an N-terminal transmembrane domain to Erd2, similar to the one in Erp1, (cytosolic YFP followed by an additional TMD) results in Golgi localization, and is partially functional for Amy-HDEL retention. Overexpression of Amy-HDEl does not result in ER re-localizatio of Erd2.

• C-tail mutations:

  • dileucine in the C-tail: mutation kills Amy-HDEL retention. The mutant localizes to the ER + Golgi.
  • deletion mutant lacking the last predicted TM domain and the cytosolic tail of Erd2. Mutant in the ER.

Major takeaways:

• Tetrapeptides KDEL and HDEL both prevent reporter protein secretion equally well in plant cells.

• Both N- and C-terminal fluorescent tagging interferes with ERD2 activity.

• N-terminal tagging of ERD2 can result in Golgi-localized fluorescent fusions as long as the ERD2 N terminus is lumenal. The lumenal N term needs to remain unobstructed for function.

• Golgi residency and biological function require a conserved dileucine motif interrupted with a nonconserved amino acid (LXL) near the Erd2 C terminus.

Summary- paper 17: Orphan quality control shapes network dynamics and gene expression

Kevin G. Mark, SriDurgaDevi Kolla, Danielle M. Garshott, Brenda Martínez-González, Christina Xu, David Akopian, Diane L. Haakonsen, Stephanie K., Michael Rapé

Biorxiv, 2022

Questions/gaps addressed:

• Accurate protein complex formation (common examples of dimerization through Zinc-fingers, BTB domains or leucine zippers) is important for the transcriptional programs that specify cell fate. How are the dynamics of these interactions accomplished?

• Transcription factors of the BTB family co-translationally form homodimers that are very stable. Transcription factors also contain activation domains that are rich in IDRs that allow recruitment of many proteins and nucleate transcription hubs which stimulate gene expression and controls cell fate decisions. How do cells establish the proper composition and dynamics of intricate transcription hubs?

• E3 ligase UBR5 was identified to play a role in preserving pluripotency in human embryonic stem cells (hESC). OE of Ubr5 in UBR5 gene in breast/ ovarian cancers s tumor growth and metastasis, and inactivating mutations in UBR5 drive mantle cell lymphoma. How does Ubr5 regulate stem cell identity and act as either a tumor suppressor or an oncogene?

Key methods:

• Binding partners of endogenously FLAG tagged UBR5 by affinity purification and CompPASS mass spectrometry. Found several subunits of the INO80 complex, DSIF complex, mitotic checkpoint complex (MCC). Confirmed most by Western blotting using antibodies against endogenous proteins.

• Test if UBR5 induces the turnover of proteins that cooperate with c-MYC, fused interactors to GFP, followed by mCherry after an internal ribosome entry site. The ratio between GFP and mCherry, was measured by fluorescence associated cell sorting (FACS). Monitored effect on stability upon siRNA against UBR5 and found that it degrades many transcriptional regulators that have links to c-MYC.

• Inhibition of p97 (NMS873), stabilized all targets. Inhibition of proteasome with MG132 also stabilized all targets. Inhibition of lysosomal function with bafilomycin A, had no effect.

• Tested specificity of UBR5 by comparing upon downregulation of other E3 ligases HUWE1, UBR4, RNF126, CHIP, WWP1, RNF168. See some overlap but UBR5 seems to be specific for transcription factors linked to c-myc.

• reconstituted the activity of this UBR5 E3 ligase in vitro. Purified UBR% from HeLa cells, and then tested in vitro ubiquitination.

• Using MCRS1 as a model substrate, tested the linkage type, and found K48 and K63 Ub both were required for UBR5 mediated ubiquitination.

• Ubiquitylation was strongly enhanced if we bypassed chain initiation by fusing ubiquitin to MCRS1, but blocked if they fused a K48 Ub, suggesting the K63 branches are added to K48 Ub

• Performed RNAseq in ΔUBR5 cells to identify genes that are upregulated. Found effects on several targets of c-MYC (both up and downregulated) and growth factors important for stem cell pluripotency.

• To test if UBR5 directly recognizes the candidate c-MYC degrons, synthesized each motif as a TAMRA-labeled peptide and incubated them with UBR5 and UBE2L3. c-MYC contains two carboxy-terminal degrons that are each sufficient to mediate recognition by UBR5.

• expressed UBR5-targets with dimerization partners and monitored effects on UBR5-dependent degradation. Complex formation stabilized all the cases shown.

Major takeaways:

• UBR5 is a nuclear HECT E3 ligase. IT synthesizes branched ubiquitin chains, requires the E1 enzyme UBA1 and the Cys-specific E2 UBE2L3.

• UBR5 targets degrons in the carboxy-terminal domain of c-MYC that become inaccessible upon formation of DNA-bound c-MYC/MAX dimers.

Summary- paper 16: The Adaptor Protein-1 μ1B Subunit Expands the Repertoire of Basolateral Sorting Signal Recognition in Epithelial Cells

Xiaoli Guo, Rafael Mattera, Xuefeng Ren, Yu Chen, Claudio Retamal, Alfonso González, Juan S. Bonifacino

Developmental Cell, 2013

Questions/gaps addressed:

• Epithelial cells are polarized into an apical side that faces the exterior or lumen of body structures and a basolateral side that contacts neighboring cells and the underlying basement membrane. Protein sorting to the basolateral plasma membrane is mediated by signals in their cytosolic tails. AP-1 adaptor plays an important role in basolateral protein sorting.

• Why do polarized epithelial cells express two isoforms of the μ1 subunit of the AP-1 clathrin adaptor complex: the ubiquitous μ1A and the epithelial-specific μ1B? What are the differences between these isoforms and how do they affect function of their respective AP-1 complexes?

• μ1A and μ1B are highly homologous (∼80% overall amino acid sequence identity in mammals).

Major hypotheses:

• Since μ1B is epithelial-specific, hypothesized that μ1B might confer on AP-1 the ability to recognize cargoes that are not efficiently recognized by μ1A.

Key methods:

• μ1A and μ1B tagged at the C termini with a 10-aa spacer (GSGSGGSGSG) followed by three copies of the hemagglutinin (HA) or Myc epitope. Transient transfection into nonpolarized Madin-Darby canine kidney (MDCK) epithelial cells. Immunostaining and confocal fluorescence microscopy showed localization of both proteins to a juxtanuclear structure characteristic of the TGN/REs, and colocalized with endogenous γ-adaptin, but not early endosomes (EEA1).

• C-terminal tag was functional based on LDLR trafficking to the basolateral surface. N-terminal tagging does not work well for the µ subunits.

• Superresolution structured illumination microscopy (SR-SIM), which has a resolution limit of ∼100 nm to look at co-localization of µ1A and µ1B. They co-localize in non-polarized cells. Saw the same thing with TIRF microscopy.

• MDCK cells grown into polarized monolayers on Transwell filters. Confocal fluorescence microscopy showed localization of endogenous γ-adaptin to a subapical compartment characteristic of the TGN/REs. μ1A-GFP and μ1B-mCherry colocalized to the same subapical compartment.

• Subcellular fractionation on 40%–60% sucrose gradients to examine the distribution of μ1A-HA and μ1B-Myc. Both cosedimented on 40%–60% sucrose gradients in association with clathrin-coated vesicles (CCVs) containing clathrin and γ-adaptin.

• Arf1 (regulated by BFA-sensitive GEFs) and Arf6 (regulated by BFA-sensitive and BFA-insensitive GEFs). Dominant-negative Arf1 mutant (Arf1-T31N) displaced both AP-1A and AP-1B complexes from TGN/REs to cytosol, Arf6 mutant (Arf6-T27N) was less efficient.

• Glutathione S-transferase (GST) pull-down assay to compare the binding of recombinant AP-1A and AP-1B core complexes to different Arfs in vitro. Both AP-1 variants bound to constitutively active forms of class I and II Arfs: Arf1, Arf4, and Arf5 and, to a lesser extent, Arf6.

• Screened the cytosolic tails of a large number of cargo proteins for interaction with μ1A and μ1B, using a yeast two-hybrid (Y2H).
  • Most tails that tested positive in this assay interacted with both μ1A and μ1B e.g., LAMP1.
  • Preferential binding to μ1A : interleukin-6 receptor α chain [IL6R-α]
  • Preferential binding to μ1B : interleukin-6 receptor β chain [IL6R-β], poliovirus receptor [PVR], low density lipoprotein receptor [LDLR].
    • The LDLR tail contains two basolateral sorting signals: a proximal signal comprising a tyrosine residue (Y828) and an acidic patch (EDE 833-835), and a distal signal comprising two tyrosine residues (Y845 and Y847) and another acidic patch (EED 856-858). All of these elements were required for interaction with μ1B.
• Pull-down using full-length LDLR tail fused to maltose-binding protein (MBP) and recombinant AP-1A and AP-1B core complexes tagged with GST, in the absence or presence of the constitutively active Arf1 Q71L mutant to test for interactions with the locked or open conformations of the AP-1 core. Pull-down with amylose beads followed by immunoblotting with antibody to GST. LDLR tail bound AP-1B ∼5-fold more avidly than AP-1A. Arf1 Q71L activated binding of both, and binding was dependent on Y845 and Y847 in the LDLR tail.

Major takeaways:

• the μ1A and μ1B subunit isoforms of AP-1 were previously proposed to specify localization of the complex to different intracellular compartments. Demonstrated that is not the case.

• Correlation between basolateral sorting and preferential interaction with μ1B for at least three cargo proteins (i.e., IL6R-β, PVR, and LDLR)

Summary- paper 15: Ferroptosis inhibition by lysosome-dependent catabolism of extracellular protein

David A. Armenta, Nouf N. Laqtom, Grace Alchemy, Wentao Dong, Danielle Morrow, Carson D. Poltorack, David A. Nathanson, Monther Abu-Remaileh, Scott J. Dixon

Cell Chemical Biology, 2022

Questions/gaps addressed:

• Cysteine is conditionally essential for survival and proliferation of many cancer cells. Cys is required for synthesis of proteins, glutathione (GSH), coenzyme A. GSH and coenzyme A are needed to prevent ferroptosis, a non-apoptotic, oxidative form of cell death. How do cancer cells manage cysteine metabolism to inhibit ferroptosis?

• GSH is a co-substrate for glutathione peroxidase 4 (GPX4), an essential enzyme that prevents ferroptosis by reducing potentially toxic membrane lipid hydroperoxides to non-toxic lipid alcohols. Mechanism of CoA in ferroptosis inhibition unclear.

• Cysteine is typically present at low abundance in the cell and in the fluid surrounding tumors in vivo. Cystine, the disulfide of cysteine is typically present outside the cell. Complete cystine starvation can induce ferroptosis in cultured cancer cell lines. How do cancer cells cope with limited extracellular amino acids, especially cystine?

• Cystine deprivation in vivo can slow tumor growth but does notcause consistent tumor regression. Are there compensatory mechanisms (besides transporter-mediated uptake of free amino acids) that limit ferroptosis in response to cystine deprivation?

• mTORC1 inhibition promotes lysosomal protein degradation. mTORC1 inhibition prevents depletion of intracellular GSH and attenuated ferroptosis in cultured cells deprived of cystine. Does this modulate cancer cell death by ferroptosis?

Major hypotheses:

• mTORC1 inhibition, together with uptake and catabolism of extracellular proteins, is sufficient to compensate for the loss of transporter-mediated cystine uptake and inhibit ferroptosis mediated cell death of cancer cells.

Key methods:

• Deprived HT-1080 fibrosarcoma cells of cystine and measured cell proliferation and cell death over time. These cells express a live cell marker, nuclear-localized mKate2 and were incubated with the dead cell dye SYTOX Green allowing monitoring cell proliferation and death to be measured simultaneously. Used inhibitors to test if the death is due to general apopotosis (pan-caspase inhibitor Q-VD-OPh), or ferroptosis (ferrostatin-1 (Fer-1)). Deprivation of cystine, but not other amino acids Met, Arg, and Leu, led to ferroptosis mediated death.

• Added serum protein albumin at a final concentration (3% [w/v]) in the media mimicking concentration in human serum, and deprived cystine in the media. Treatment with the ATP-competitive mTOR inhibitor INK128 (or other mTOR inhibitors Torin 1 and rapamycin or shRNA targeting the mTORC1 subunit RPTOR) was sufficient to attenuate cystine deprivation-induced cell death.

• Saw similar effects of combination treatment in cystine-deprived cancerous cell lines: H1299 and H23Cas9 non-small cell lung carcinoma cells and U-2 OS osteosarcoma, A375 melanoma, T98G glioblastoma, and PaTu 8988T pancreatic adenocarcinoma cell lines.

• Albumin + INK128 treatment protected equally well against ferroptosis in control and ATG7 KO PaTu 8988T cell lines, suggesting that autophagy does not mediate this effect. Treatment did not increase expression of the anti-ferroptotic proteins GPX4 or FSP1, or decrease intracellular iron levels, as assessed by expression of the transferrin receptor (TFRC) or IRP2, which are induced by iron starvation.

• Monitored membrane lipid peroxidation in cystine-deprived cells using the sensor C11 BODIPY 581/591. Reduced oxidized signal when also treated with albumin. Albumin treatment alone weakly inhibited ferroptosis in response to low (but not high) doses of the GPX4 inhibitors FIN56 and ML162, Albumin can possibly act as a direct antioxidant to some extent. The effect was a lot more when combined with cystine deprivation.

• Albumin + INK128 treatment + Inhibition of lysosomal proteases with deacidifying agents (e.g., chloroquine) or using a cocktail of the protease inhibitors pepstatin A, leupeptin- cells died again (A375, H1299, and T98G cancer cells, but no effect in U-2 OS, unclear why). Also lysosomal function by microscopy using fluorescent probe DQ-BSA.

• Albumin + INK128 treatment + Cystine deprivation + treatment with cathepsin inhibitors: CTSB inhibitor CA074 methyl ester but not the cathepsin L and S inhibitor R11-OEt: led to DQ-BSA fluorescence quenching, i.e. Cathepsin B CTSB predominantly responsible for albumin breakdown in lysosome. No defect in endocytosis or internalization of tetramethylrhodamine (TMR)-dextran. Same effect observed upon genetic disruption of CTSB, CTSL, CTSD.

• Examined the role pf Cystinosin (CTNS) in exporting cystine produced by Albumin breakdown by CTSB from the lysosomal lumen to the cytosol. Generated CTNS KO HT-1080 cells. Albumin + INK128 treatment + Cystine deprivation + CTNSKO cells, cells died from ferroptosis and had high lipid peroxidation.

• Established HT-1080 spheroids to mimic tumor microenvironment over 3 days in ultra-low adherence vessels. vehicle + system xc− inhibitor erastin2 cells died from ferroptosis. Albumin + system xc− inhibitor erastin2 cells lived.

Major takeaways:

• Endocytosis and lysosomal degradation of extracellular albumin, provides enough cysteine, and hence GSH to cancer cells to prevent death by ferroptosis, even in the absence of cystine uptake from the environment. This protective mechanism requires the lysosomal enzyme cathepsin B and the lysosomal cystine exporter cystinosin.

Summary- paper 14: Calcium levels in the Golgi complex regulate clustering and apical sorting of GPI-APs in polarized epithelial cells

Stéphanie Lebreton, Simona Paladino, Dandan Liu, Maria Nitti, Julia von Blume, Paolo Pinton, Chiara Zurzolo

PNAS, 2021

Questions/gaps addressed:

• Glycosylphosphatidylinositol-anchored proteins (GPI-APs) are lipid-associated luminal secretory cargos. Distinct clustering mechanisms drive their sorting out of the Golgi apparatus to the cell surface in polarized cells and non-polarized cells. What are the mechanisms of GPI-AP apical sorting in the Golgi apparatus?

• Golgi exit of various cargos is altered in cells treated with drugs either depolymerizing/ stabilizing actin filaments/ mutants affecting actin function. High calcium levels in the Golgi are important for sorting of some secreted soluble proteins. Possible role of the actin cytoskeleton and of calcium levels in the Golgi?

Major hypotheses:

• The amount of calcium in the Golgi lumen regulates the formation of GPI-AP homoclusters in polarized epithelial cells.

Key methods:

• Polarized MDCK (Madin-Darby Canine Kidney) cells treated with ionomycin to deplete calcium ions from the Golgi (preferentially depletes Golgi luminal calcium) amd monitored effect on localization of a model apical GPI-AP, GFP-FR. Observed the treatment resulted in reduced accumulation of HMW or multimeric complexes of GFP-FR.

• Measured the calcium concentration in the Golgi by using a calcium-sensitive photoprotein (Golgi-aequorin chimera). Found calcium levels in the Golgi apparatus of polarized MDCK cells are slightly higher compared to the Golgi of nonpolarized MDCK cells.

• Effect of two groups of phosphorylation-type calcium pumps, SERCA and SPCA1. SPCA1 protein levels increase with the establishment of polarity. shRNA KD of SPCA1 leads to reduction in the concentration of calcium in the Golgi using the calcium sensor, and reduced the homoclustering of GFP-FR.

• Transport kinetics of GFP-FR using time-lapse confocal microscopy upon SPCA1 KD to look at Golgi export using temperature block assays. KD affected both the Golgi exit and the apical sorting. Cab45, a Golgi (TGN) luminal protein, oligomerizes upon calcium binding and required for protein sorting. KD of Cab45 lead to missorting of GFP-FR to the basolateral membrane. Observed similar effects on endogenous GPI-AP, PLAP (placental alkaline phosphatase).

• Fluorescent-conjugated bacterial toxin aerolysin, which binds with high affinity GPI-APs to monitor surface distribution of endogenous GPI-APs upon Cab45 KD, observed missorting to basolateral surface.

• Treatment with latrunculin A to perturb actin cytoskeleton and monitored GFP-FR localization and clustering, no effect of actin perturbation on the oligomeric status.

Major takeaways:

• While high cholesterol concentration is necessary for the clustering of GPI-APs, this is not sufficient for their apical sorting.

• Golgi organization of GPI-APs is drastically perturbed upon calcium depletion and that the amount of calcium in the Golgi cisternae is critical for the formation of GPI-AP homoclusters.

• Down-regulation of the TGN calcium/manganese pump, SPCA1 (secretory pathway Ca(2+)-ATPase pump type 1) or Cab45, a calcium-binding luminal Golgi resident protein, impairs the oligomerization of GPI-APs in the Golgi complex and leads to their missorting to the basolateral surface but does not affect apical or basolateral transmembrane proteins, or basolateral GPI-APs such as PrP.

• Levels of SPCA1 increase during polarization, while levels of Cab45 decrease (but oligomerization increases) during polarization.

• Calcium-binding proteins, such as calmodulin, are thought to mediate fusion between yeast vacuoles, the later steps of fusion vesicle trafficking and endosome fusion. Do Golgi luminal calcium-binding proteins, Cab45 and p54/NEFA (yeast SSP120), work in similar ways?

• The calcium-dependent mechanism for GPI-AP sorting is dominant over dependence on Golgi cholesterol levels, apical sorting of GPI-APs. What is the interplay between Golgi cholesterol levels and Calcium levels?

Summary- paper 13: Gene dosage screens in yeast reveal core signalling pathways controlling heat adaptation

Cosimo Jann, Andreas Johansson, Justin D. Smith, Leopold Parts, Lars M. Steinmetz

Biorxiv, 2020

Questions/gaps addressed:

• Heat shock induces significant transcriptional and transational changes in yeast and higher organisms but bulk of heat-induced changes seem dispensable. What mechanisms are essential for cell survival?

Key methods:

• Inducible CRISPRi/a systems using a catalytically inactive Cas9 nuclease fused to transcriptional repression or activation domains to modulate levels of 129 protein kinases and 161 transcription factors in yeast (6 sgRNAs/gene), to screen for effects on cellular fitness at temperatures 23°C, 30°C and 38°C.

• Employed the Streptococcus pyogenes Cas9 fused to the human Mxi1 repressor which both evolved to operate around 37°C. Tet-inducible dCas9-MxiI and dCas9-nGal4-VP64 plasmids (AddGene #73796 and #71128).

• Chemically synthesized gRNA oligonucleotide libraries (CustomArray, Inc. (GenScript)) amplified by PCR and integrated into pRS416 dCas9-Mxi1 plasmid via Gibson Assembly with 30bp homology regions, or ligation with T4 DNA Ligase. - R code for screen analysis on Github

• Validate repression/ activation from anhydrotetracycline (ATc)-inducible gRNA agaisnt Hsf1 transcription factor and monitor effect on growth rate. Quantify Hsf1 function with a truncated promoter of the SSA1 HSP70 gene driving GFP reporter, and FACS to monitor inheritibility.

Major takeaways:

• CRISPRi efficiency depends on the GC content and secondary structure of gRNAs. The optimal range forgRNA target locus is between TSS-150 to TSS+25 nucleotides, with minor variation between target strands. Effective repression observed in CRISPRi strains over time and all temperatures.

• Multiple pathways (HSR, HOG, UPR, cell cycle) controlling HSR activity that affect chaperone stability and expression levels needed for thermotolerance.

Summary- paper 12: A CRISPRi/a screening platform to study cellular nutrient transport in diverse microenvironments

Christopher Chidley, Alicia M. Darnell, Benjamin L. Gaudio, Evan C. Lien, Anna M. Barbeau, Matthew G. Vander Heiden, Peter K. Sorger

Biorxiv, 2023

Questions/gaps addressed:

• Aggressive growth of cancerous cells means dependence on teh environment for supplies of non-essential amino acids. This includes an increase in the rate of nutrient uptake, by upregulation of SLC transporters and a rewiring of intracellular metabolism.

• Which SLC transporters are more important for tumor growth, and can they used as therapeutic targets.

Major hypotheses:

• Key SLC and ABC transporters play an essential role in proliferation of cancerous cells.

Key methods:

• CRISPR interference (CRISPRi) and activation (CRISPRa) growth-based pooled screens (10 sgRNAs/gene and 730 non-targeting controls) to systematically interrogate the individual contributions to nutrient transport and growth by 489 annotated members of the SLC and ABC transporters in human cells.

• Identified all amino acids in RPMI-1640 medium whose absence limited K562 cell proliferation by performing proliferation assays with a series of single amino acid dropout RPMI media to which amino acids were added back to between 0.1% and 100% of their normal levels.

• Performed pooled CRISPRi/a screens at amino acid concentrations that reduced K562 cell proliferation by 50% for all 13 growth-limiting amino acids to identify transporter perturbations that caused either a decrease or increase in proliferation. Three rounds of low amino acid exposure lasting for four days with daily media changes followed by one recovery day in complete medium, enrichment analysis between starved and complete RPMI conditions.

• Validated phenotypes in PAA-RPMI medium formulated to contain all amino acids as well as five other known SLC7 family substrates (citrulline, ornithine, creatine, creatinine, and carnitine) at levels found in human plasma.

• Mass spectrometry-based transport assays to follow amino acid transport at the plasma membrane. Quantified amino acid import by incubating K562 cells in medium containing amino acids labeled with heavy isotopes and measuring intracellular isotope accumulation by gas chromatography-mass spectrometry (GC-MS)

• Tested the effect of two ferroptosis inducers: erastin, which inhibits SLC7A11, and RSL3, which inhibits glutathione peroxidase 4 (GPX4).

• Compared transporter essentiality in: (1) RPMI at high confluence, (2) RPMI with regular fetal bovine serum (FBS) replaced with FBS that was dialyzed to remove small molecules (dFBS), (3) Dulbecco’s Modified Eagle Medium (DMEM), an alternative synthetic medium, and (4) in different cell lines- K562 vs A375 (BRAFV600E melanoma)

Major takeaways:

• Removal of Asn, Asp, Glu, Gly, and Pro from the RPMI-1640 medium has no effect on the proliferation of K562 cells. Only these five amino acids are also net secreted by K562s. Cystine (oxidized cysteine more abundant in culture medium)deprivation induces significant cell death (ferroptosis?). A reduction in media levels of 13 of the 19 amino acids found in RPMI decreased proliferation.

• Strong CRISPRi hits were also hits in the corresponding CRISPRa screens. CRISPRi/a combination screens also allow identification of genes expressed in a particular cell line.

• RPMI which contains many amino acids at 0.4× to 10× the levels found in human plasma but only trace amounts of alanine, cysteine.

• Serotonin (5-HT) and transporter SLC6A4 act as an endogenous antioxidant to protect cells from ferroptosis.

• Specific transporters differentially affected depending on the influence of the environment on metabolism, and differences in metabolism between cell lines.

Summary- paper 11: MitoStores: chaperone-controlled protein granules store mitochondrial precursors in the cytosol

Lena Krämer, Niko Dalheimer, Markus Räschle, Zuzana Storchová, Jan Pielage, Felix Boos, Johannes M Herrmann

The EMBO Journal, 2023

Questions/gaps addressed:

• Most mitochondrial proteins are synthesized and folded in the cytosol, then imported into mitochondria by translocases in the mito membranes. Accumulation of nonimported precursor proteins in the cytosol upregulates Rpn4 (transcription factor)-dependent Unfolded Protein Response activated by mistargeting of proteins (UPRam). How does the cell handle the stress of mito protein accumulation in the cytosol when UPRam upregulation is blocked?

Key methods:

• Overexpression cloggers- mito precursors (b2Δ-DHFR or b2-DHFR) and monitor effect of Rpn4 deletion on growth.

• Comparison of total cellular proteomes of wild-type and rpn4∆ cells with/without clogger induction by TMT MS.

• Affinity purification with nanotraps-MS from yeast lysates (+/- clogger) to identify the content of Hsp104-GFP-bound structures.

• Split-GFP approach (first 10 beta sheets of superfolder GFP in mitochondria (Oxa1-GFP1-10)) and strand 11 of GFP on candidates to confirm location.

Major takeaways:

• Overexpression of slowly imported mito precursors (b2Δ-DHFR or b2-DHFR) called cloggers, causes growth arrest. OE in rpn4∆ mutant no longer arrested for growth. Rpn4 dependent toxicity was specific for mito precursors, no effect on OE of cytosolic aggregation-prone proteins eg., polyglutamine protein Q97-GFP, luciferase(R188Q R261Q), or Ubc9 Y68L.

• OE of clogger in rpn4∆ cells: proteasome subunits downregulated, cytosolic chaperone Hsp104 (disaggregase) and Hsp42 (small heat shock protein) upregulated. rpn4∆ hsp42∆ hsp104∆ triple mutant cannot grow on respiratory media/nonfermentable carbon sources.

• Hsp104-GFP accumulates in cytosolic puncta upon clogger OE. IP-MS of Hsp104-GFP structures from cells identified multiple mitochondrial proteins in the IP. Authors call these structure ‘MitoStores’. Found Pdb1, Atp14, Tdh2, Cox8, Cox13, Mrp3 as stronger Hsp104 interactors in rpn4∆ bkg.

Summary- paper 10: A dual sgRNA library design to probe genetic modifiers using genome-wide CRISPRi screens

Alina Guna, Katharine R. Page, Joseph R. Replogle, Theodore K. Esantsi, Maxine L. Wang, Jonathan S. Weissman, Rebecca M. Voorhees

Biorxiv, 2023

Questions/gaps addressed:

• Establishing a pipeline that permits querying epistatic relationships with a variety of phenotypic readouts in any cells expressing the CRISPRi machinery.

Key methods:

• CRISPR interference (CRISPRi) based approach with a flexible dual-sgRNA CRISPRi-v2 library (Addgene #83969; 5 sgRNA/gene library targeting human protein-coding genes) compatible with Fluorescence Activated Cell Sorting (FACS)-based reporter screens.

• Construct and deliver a library containing a fixed pre-determined guide, genetic ‘anchor point’, and a second randomized CRISPRi guide from a single lentiviral backbone at scale.

• Two parallel screens, one with a ‘non-targeted contol’ and the other target gene as ‘ achor point’.

• FACS-based approach, adapted a fluorescent split GFP reporter system to look at relationships between the parallel pathways specifically to query insertion into the ER. Constitutively expressed the first 10 β-strands of GFP (GFP1-10) in the ER lumen and appended the 11th β -strand onto the C-terminal of the endogenous sequence of SEC61β (SEC61β-GFP11). Engineered K562 cells to stably express ER GFP1-10 and the dCas9-KRAB(Kox1) CRISPRi machinery. Inducible promoter, integrated the SEC61β-GFP11 reporter alongside a normalization marker (RFP) separated by a viral 2A sequence, allows to use the GFP:RFP ratio to differentiate transcriptional and translational differences.

Major takeaways:

• Approach eliminates the need to create and characterize a knock-out line for a particular gene of interest, to set up epistatsic genetic interation screens.

• Construction of new libraries is easy/fast, with a two-step cloning process,uses existing sequencing and analysis pipelines.

• Caveat: addition of a second guide delivered on the same plasmid diminishes the efficiency of the fixed guide. Alternative using Zim3-Cas9 effector system, with stronger on-target knockdown compared to KOX1-Cas9 while maintaining minimal non-specific genome-wide effects.

Summary- paper 9: A genome-wide CRISPR interference screen using an engineered trafficking biosensor reveals a role for RME-8 in opioid receptor regulation

Brandon Novy, Hayden Adoff, Monica De Maria, Martin Kampmann, Nikoleta G. Tsvetanova, Mark von Zastrow, Braden Lobingier

Biorxiv, 2022

Questions/gaps addressed:

• Post agonist-stimulated-endocytosis trafficking of GPCRs from the cell surface can result in receptor degradation in the lysosome or recycling back to the plasma membrane. The fate depends on a combination of sorting motifs and dynamic post-translation modifications. What are the machineries involved that make these sorting decisions?

Key methods:

• Chemical functional genomics FACS-based pooled screen using APEX2 driven activation of fluorogenic substrate Amplex UltraRed as a read-out to monitor GPCR trafficking.

• Delta opioid receptor (DOR) fused to APEX2 at the cytsoplasmic C-terminus (DOR-APEX2). When in the cytosol, APEX2 can convert the non-fluorescent substrate Amplex UltraRed (AUR) to a fluorescent resorufin-like molecule in the presence of hydrogen peroxide.

• Upon agonist (DADLE) addition, DOR-APEX2 is delivered to the MVB and lysosome lumen, leading to DOR-APEX2 degrdation by proteases at low pH, resulting in loss of fluoresence in lysates. Lysate APEX2/AUR reaction can be quenched with 100-fold molar excess of a competitive substrate (sodium L-ascorbate).

• Adapting the assay to intact cells: Applying a transient pulse of sodium azide to quench the reaction and bovine serum albumin to scavenge excess dye. RNAi against Arrestin3, blocks DOR endocytosis, resulting in stabilization of fluorescence signal monitored by red laser (APC channel: Ex: 633 nm; Filter: 670/30 nm)

• Genome-wide CRISPR interference (CRISPRi) dCas9 based screen and second generation libraries encompassing the entire human genome. HEK293 cell line stably expressing dCas9-KRAB and pUBC:DOR-APEX2. Seven sub-libraries consisting of 5 sgRNAs/gene and controls. Estimated sub-library coverage of 500-fold. Following APEX2/AUR reaction, cells sorted to isolate the high and low fluorescence. Genes identified by next generation sequencing.

Major takeaways:

• The activity of APEX2 that is sensitive to mildly acidic conditions.

• Identified hits of genes functioning along the endocytic and secretory pathway. Followed-up on some hits that were not known to function with DOR: DNAJC13/ RME-8 (strongest stabilization of fluorescence), WDR91, and SNX24.

• RME-8 (no yeast homolog), conserved from worms to humans, peripherally associates with EEA1-positive endosomes. RME-8 and endocytosed DOR2 were adjacent not overlapping. RME-8 signal in between DOR2 and VPS35. RME-8 is enriched at sites of endosomal tubulation/recycling?

• No defect in lysosomal delivery upon KD of other components of the recycling pathway, VPS35 or FAM21A/C. RME-8 appears to be anew player modulating the lysosomal delivery of DOR-APEX2. Mechanism unclear. Cool genetic screen approach.

• Flexibility of the APEX2 enzyme to mediate oxidation of different chemical substrates allows switching between proximity labeling and fluorogenic activation in the same cell line. Cool strategy!

Summary- paper 8: SLC25A39 is necessary for mitochondrial glutathione import in mammalian cells

Ying Wang, Frederick S. Yen, Xiphias Ge Zhu, Rebecca C. Timson, Ross Weber, Changrui Xing, Yuyang Liu, Benjamin Allwein, Hanzhi Luo, Hsi-Wen Yeh, Søren Heissel, Gokhan Unlu, Eric R. Gamazon, Michael G. Kharas, Richard Hite & Kıvanç Birsoy

Nature, 2021

Questions/gaps addressed:

• Mitochondria (highly redox-active), contain 10–15% of total cellular Glutathione (GSH), but lack GSH biosynthetic machinery. How does GSH get inside the mitochondria?

Major hypotheses:

• Since GSH is negatively charged under physiological conditions, there must be a dedicated GSH transporter in the mitochondrial membrane.

Key methods:

• Mito-IP from HeLa cells expressing a mitochondrial tag (3×HA–OMP25–mCherry) grown in standard medium or treated with a GSH synthesis inhibitor, buthionine sulfoximine (BSO), followed by proteomics to identify proteins with altered abundance.

• Metabolomics on mito-IPs to profile changes in mitochondrial metabolites in WT vs SLC25A39-KO cell lines.

• in vitro GSH transport assay using isotope-labeled GSH (GSH-(glycine-13C2,15N))-uptake assays using mitochondria isolated from WT or SLC25A39-KO cells.

• Genetic interaction analysis: CRISPR–Cas9-based genetic screens in both Jurkat and HEK 293T cells using a metabolism-focused sgRNA library to screen for genes essential for cellular proliferation in the absence of SLC25A39.

• Mitochondrially targeted GshF (Streptococcus thermophilus bifunctional enzyme with both glutamate–cysteine ligase and GSH synthetase) to complement mitochondrial GSH loss.

Major takeaways:

• Identified SLC25A39 (uncharacterized mitochondrial membrane small molecule transporter) as highly upregulated proteins upon GSH depletion with BSO, or upon loss of GCLC (rate-limiting enzyme of GSH synthesis).

• Observed largest reductions in the levels of GSH or GSH disulfide in mitochondria (Mito-IP) of SLC25A39-knockout cells by metabolomics.

• SLC25A39 dependent mitochondrial uptake specific for GSH (observed no uptake of GSSG or GSH disulfide) in in-vitro uptake assay. Mutating K329A or D226A (conserved residues as potential substrate-binding residues, modeled based on the structure of the bovine ADP/ATP transporter) abolished GSH uptake. Expression of mito-GshF complemented mito GSH levels in SLC25A39-KO cells.

• Redundant role of SLC25A40 and SLC25A39 in mitochondrial GSH uptake. Identified SLC25A40 (mitochondrial SLC25A family transporter) with the highest sequence homology to SLC25A39, in the genetic interaction screen. Double mutants unable to proliferate under standard growth conditions. (But did not identify SLC25A40 levels upregulated upon GSH inhibition?)

• SLC25A40 and SLC25A39 are highly conserved: yeast Mtm1, Drosophila Shawn.

• Downregulation of mitochondrial translation and iron–sulfur cluster-containing proteins in unbiased proteomics on SLC25A39/40 double-ko Jurkat cells. Slc25a39-KO mice embryonically lethal E13.5, and embryos appeared pale, severely anaemic. Importance of GSH import during erythropoiesis? Effect on iron–sulfur cluster biogenesis due to GSH’s role as a cofactor for mitochondrial glutaredoxins?

Summary- paper 7: The yeast endocytic early/sorting compartment exists as an independent sub-compartment within the trans-Golgi network

Junko Y. Toshima, Ayana Tsukahara, Makoto Nagano, Takuro Tojima, Daria E. Siekhaus, Akihiko Nakano, Jiro Toshima

Biorxiv, 2022

Questions/gaps addressed:

• How and where do uncoated endocytic vesicles get delivered to the early/sorting endosome compartment in yeast? i.e. What is the first destination of endocytic cargos in yeast? Which Q-SNARE (target SNARE) engages with the vesicles (Tlg2?)?

• Conflicting observations in literature- (1) Sec7-residing TGN (trans-Golgi network) is the first endocytic compartment; (2) two endosomal compartments in yeast- one with Rab5/Vps21, other with Rab7/Ypt7; (3) endocytosed cargos rarely transported to Sec7-TGN. Why the inconsistencies? Limitation in imaging technologies?

Key methods:

• Simultaneous triple-color and 4D (3D plus time) super-resolution confocal live imaging microscopy (SCLIM) to monitor cargo trafficking in yeast

Major takeaways:

• Localization dynamics of Tlg2 and Sec7 at the TGN are distinct. Tlg2 appears at the TGN earlier than Sec7 and disappears before Sec7 from the TGN.

• Localization of pHrode Red-α-factor (endocytic cargo) overlaps more with Tlg2 5 min post endocytosis than Sec7. Tlg2 appears to localize with the cargo throughout the early-to-late stage endosomes.

• Vps21p partially localizes around the TGN (overlap with Sec7) as well as in endosomal compartments (overlap with Tlg2). Tlg2 is transported to the PVC (prevacuolar compartment) via the Vps21-residing compartment without passing through the Sec7-TGN. How does Tlg2 get back to the early-TGN?

• γ-adaptin ear containing, Arf binding protein (GGA) adaptors, Gga1 and Gga2; and epsin-related proteins, Ent3/5 are required for delivery of α-factor from TGN to PVC. AP1 adaptor not needed. GGAs arrive at the TGN earlier than AP1.

• Gga1/2 important for delivery of Tlg2 compartment to endosome (signal for Gga- Ub on cargos destined for lysosomal degradation?). Ent3/5 important for Vps9 (GEF for Vps21) delivery from TGN to endosome. Nice model figure!

Summary- paper 6: Dynamic quality control machinery that operates across compartmental borders mediates the degradation of mammalian nuclear membrane proteins

Pei-Ling Tsai, Christopher J.F.Cameron, Maria Fernanda Forni, Renee R.Wasko, Brigitte S.Naughton, Valerie Horsley, Mark B.Gerstein, Christian Schlieker

Cell Reports, 2022

Questions/gaps addressed:

• What are the molecular mechanisms responsible for protein folding and turnover in the inner nuclear membrane (INM)? ER is contiguous with the outer nuclear membrane (ONM), not INM, which E3 ligases surveil the INM?

Major hypotheses:

• ONM/INM diffusion barrier necessitates that a dedicated INM-resident machinery exists to perform a ERAD-like role.

Key methods:

• LBR1600 Lamin B receptor (LBR) mutant, responsible for Pelger-Huët anomaly and Greenberg skeletal dysplasiais metabolically unstable/ short lived, serves as a INM PQC model substrate.

• split-GFP system to monitor the status of mutant LBR1600 localization (last beta strand of GFP (S11) fused to its C terminus) and NLS-GFP1-10. Both expressed froma Tet-inducible retroviral system.

• GFP fluorescence stabilization as a read-out for CRISPR-Cas9-based screen (Brunello KO library- 4 sgRNAs/gene against 19,114 genes; 1,000 sgRNA controls) to identifying LBR1600 turnover machinery. Cells with sgRNAs and stabilized GFP fluorescence sorted and enriched (2 rounds), and sgRNA frequency analyzed. Proof of principle stabilization upon treatment with VCP INHIBITOR, CB-5083.

Major takeaways:

• LBR1600 turnover is stabilized in KO of E2 ubiquitin-conjugating enzymes Ube2G2 and Ube2D3, as well as the E3 Ub ligases RNF5 and HRD1/SYVN1 (confirmed by rescue and catalytic inactive mutants), and uncharacterized membrane protein TMEM33. All resulted in reduced polyubiquitination, predominandtly K48 linkage. VCP was not identified as a hit in the screen.

• RNF5-HA (CRISPR/Cas12a-mediated PCR tagging) immunoprecipitates LBR1600-GFP11. TMEM33-HA pulls down FLAG-RNF5 as well as LBR1600.

• TMEM33 works upstream of RNF5 in the same pathway for LBR1600 turnover, HRD1 acts via an independent pathway. OE of RNF5 in hrd1-KO leads to LBR1600 degradation. What about OE of HRD1 in rnf5-KO cells? Is either sufficient?

• TMEM33 is required for RNF5 protein stability- act as a chaperone for RNF5 or binding might prevent RNF5 auto-ubiquitination?

• RNF5 localizes to the ER and INM- authors propose that RNF5 is small (20kD) and may easily pass the ER-INM diffusion barrier, hence accessign INM substrates. If Hrd1 cannot pass this barrier- how then does it have an effect on LBR1600 turnover?

Summary- paper 5: Organelle proteomic profiling reveals lysosomal heterogeneity in association with longevity

Yong Yu, Shihong M. Gao, Youchen Guan, Pei-Wen Hu, Qinghao Zhang, Jiaming Liu, Bentian Jing, Qian Zhao, David M Sabatini, Monther Abu-Remaileh, Sung Yun Jung, Meng C. Wang

Biorxiv, 2022

Questions/gaps addressed:

• Lysosomes are the key platform to modulate the activities of mTORC1 and AMPK signaling, both of which regulate longevity. How does the composition of lysosomes change as their activities modulate or vary with ageing?

Key methods:

• Rapid Lysosome IP, using 3HA tagged TMEM192 transgenic C.elegans (in whole worms, or tissue specific expression in epidermis, muscle, intestine, and neurons),followed by large-scale proteomic profiling.

• Comparison of lysosome-enriched proteome with LAMP1/LMP-1 (LMP1-RFP-3HA) Lyso-Tag and the one with Cystinosin/CTNS-1 (CTNS-RFP-3HA) Lyso-Tag.

• in vitro staining of isolated lysosomes with Lysotracker to monitor the intactness of isolated lysosomes

• RNAi screen in worms based on LysoSensor fluorescence intensity.

Major takeaways:

• LAMP1 LysoIP: 216 lysosome-enriched candidates identified, 178 candidates have mammalian homologs. No CTNS in this set.

• LAMP1 LysoIP in long lived worms: lipl-4 (lysosomal acid lipase) and daf-2 (insulin/IGF-1 receptor): significant number (vATPase, autophagy, mTOR, AMPK related) enriched in mutant (lipl-4) vs wt. Also Nucleoporins (lysosomes also more perinuclear in lipl-4). More mature lysosomes accumulate in mutants?

• CTNS LysoIP (supplement 6) vs LAMP1 LysoIP: 95 common proteins, more mature lysosomes in CTNS IP.

• RNAi KD of CTNS and LAMP1 common (95) proteins, followed by lysosomal staining by LysoSensor. vATPAse mutant reduced lysosomal number, increased size. One new effector- TMEM144- predicted carbohydrate transporter, mutant increases lyso pH.

• Heterogeneity of lysosomal protein composition associated with lysosomal status, tissue specificity, and organism longevity.

Summary- paper 4: A microbial transporter of the dietary antioxidant ergothioneine

Daniel G. Dumitrescu, Elizabeth M. Gordon, Yekaterina Kovalyova, Anna B. Seminara, Brianna Duncan-Lowey, Emily R. Forster, Wen Zhou, Carmen J. Booth, Aimee Shen ,Philip J. Kranzusch, Stavroula K. Hatzios

Cell, 2022

Questions/gaps addressed:

• Low-molecular-weight (LMW) thiols such as Glutathione are key to maintaining redox balance. Helicobacter pylori (gastric pathogen) lacks enzymes to synthesize LMW thiols. How does it maintain intracellular redox homeostasis without producing known LMW thiols, despite persistent exposure to ROS in the host?

Major hypotheses:

• H. pylori must have mechanisms such as a transporter to obtain LMW thiols from the host.

Key methods:

• reactivity-guided metabolomics screen to identify LMW thiols in H. pylori, wherein bacterial cell extracts were treated with the thiol-alkylating agent monobromobimane (mBBr)

• crystal structure of EGT bound to EgtU transporter

Major takeaways:

• Metabolomic screen identified ergothioneine (EGT) is present in H. pylori. EGT is a sulfur-containing derivative of histidine, stable under physiological conditions, only synthesized by bacteria (no H pylori) and fungi (so come via diet).

• H. pylori also lacks the mammalian EGT transporter (OCTN1) ortholog.

• EGT structure resembles bacterial osmolyte glycine betaine. Found that betaine transporters EgtUV can also import EGT.

• Presence of these EgtUV transporters provides H. pylori competive colonization advantage. These transporters are conserved in GI microbes.

Summary- paper 3: Lipid and protein content profiling of isolated native autophagic vesicles

Daniel Schmitt, Süleyman Bozkurt, Pascale Henning-Domres, Heike Huesmann, Stefan Eimer, Laura Bindila, Christian Behrends, Emily Boyle, Florian Wilfling, Georg Tascher, Christian Münch, Christian Behl, Andreas Kern

EMBO Reports, 2022

Questions/gaps addressed:

• What is the impact of different autophagy conditions on the exact Phospholipid and cargo contents of autophagosomes/autophagic vesicles?

• Optimize a method for the isolation of unmanipulated autophagic vesicles at large quantities for rapid lipid and cargo profiling.

Key methods:

• anti-Atg8 antibody-based FACS mediated isolation approach to purify intact native autophagic vesicles from bafilomycin A1 (accumulates autophagic vesicles) treated HeLa cell homogenates.

• negative stain electron microscopy to visualize intact vesicles, and proteinase K digestion assay for p62/SQSTM1 to evaluate intactness

• quantitative lipidomics and proteomics analyses to identify PLs and cargo proteins of autophagic vesicles

Major takeaways:

• Antibody binding in cell homogenates and FACS enrichment based pipeline can isolate intact autophagic vesicles

• Size evaluation of isolated vesicles: diameters range from 340 to 1,150 nm

• phosphatidylcholine (PC) enriched: accounts for 42% of all detected PLs in autophagic vesicles. Also contain sphingomyelin (SM). Phosphatidylglycerol (PG) appears to be excluded.

• Protein composition of autophagic vesicles (dataset EV1): identified >4000 proteins. Composition varied more upon proteasome inhibition than starvation.

• Approach does not distinguish between autophagosomes and autolysosomes.

Summary- paper 2: Bidirectional promoter activity from expression cassettes can drive off-target repression of neighboring gene translation

Emily Nicole Powers, Charlene Chan, Ella Doron-Mandel, Lidia Llacsahuanga Allcca, Jenny Kim Kim, Marko Jovanovic, Gloria Ann Brar

eLife, 2022

Questions/gaps addressed:

• Selection-cassette based genetic engineering (aka Longtine strategy) is a routine strategy in yeast- are there negative consequences associated with this approach?

• Do selection-mediated editing strategies in yeast lead to transcriptional interference by long undecoded transcript isoforms (LUTIs) of neighboring genes (neighboring gene effect or NGE)? How can this effect be avoided?

Major hypotheses:

• Insertion of expression cassettes commonly used to edit the genomes of yeast cells, can induce the repression of neighboring genes though synthetic and constitutive LUTI-based repression

Key methods:

• Ribosome profiling and mRNA-sequencing (mRNA-seq)

• Label-free mass spec on fractionated polysomes and hierarchical clustering to assess global differences in polysome composition

• Rapid amplification of cDNA ends (5′RACE)

Major takeaways:

• Stable cassette-driven divergent transcripts (in most selection marker cassettes TRP1/Kan/Hyg/Nat) were detected in more than 30% of cassette-inserted loci

• Bidirectional promoter activity from the TEF promoter most likely cause of LUTI-based repression​

• Bidirectional promoters are very common, including pTEF in yeast, and the CMV, eEF1α, and SV40 promoters in mammalian systems

• Rrp6/exosome-mediated degradation can mask the divergent transcripts produced from cassette insertion in many cases​

• Placement of strong transcription terminator sequences flanking both ends of the resistance cassette prevents neighboring gene disruption​, provide modified Longtine based cassettes with these changes.

Summary- paper 1: Dynamic metabolome profiling uncovers potential TOR signaling genes

Stella Reichling, Peter F Doubleday, Tomas Germade, Ariane Bergmann, Robbie Loewith, Uwe Sauer, Duncan Holbrook-Smith

eLife, 2023

Questions/gaps addressed:

• what is the metabolomic response of TOR related mutants when exposed to Rapamycin? Can these responses be used as signature to identify new genes in the pathway? Can this approach be more generalized for unknown pathways?

Major hypotheses:

• Metabolomic signatures can be used to cluster and functional chracterization of unknown genes

Key methods:

• High-throughput metabolomics by flow injection analysis chromatography free mass spectrometry (FIA-MS): provides broad coverage at < 1min/sample

• Pearson correlation between metabolome profiles of TOR vs receptor collection. Manhattan distances between profiles of correlation followed by hierarchical clustering. False positive rate set to 0.2

• Label free quantitative (LFQ) proteomics to compare proteomic changes.

Major takeaways:

• Dynamic high-throughput metabolomics of 164 non-essential yeast deletion mutants (include 80 known TOR related, and 80 receptor related- supplement 1)- found 3 candidates as positive regulators of TOR signaling (YML079W/CFF1; Bck1; Cla4)

• Cool and interesting approach with potential. Mechanistic insight or if the effects of the candidates are direct or indirect is unclear.

• Data visualization web resource

One of the goals for this year is to commit to reading more scientific literature. To avoid letting it slip down the priority list, I’ll try to document the paper and add a brief summary (in the following format) on the blog everyday.

• Questions/gaps addressed:

• Major hypotheses:

• Key methods:

• Major takeaways:

Sardana lab website is up and running!

Can’t wait to get the science started!

The site was generated using Jekyll Bootstrap and borrows its template from the Drummond lab webpage. The site’s source code is freely available on GitHub. Drummond lab site was in turn heavily influenced and templated on Bedford lab’s webpage.

Research Paper Highlight: DNA origami signposts for identifying proteins on cell membranes by electron cryotomography

Emma Silvester, Benjamin Vollmer, Vojtěch Pražák, Daven Vasishtan, Emily A Machala, Catheryne Whittle, Susan Black, Jonathan Bath, Andrew J Turberfield, Kay Grünewald, Lindsay A Baker

Cell 2021 Feb 18;184(4):1110-1121.e16. doi: 10.1016/j.cell.2021.01.033.

For Part 3 of the Structuromics series I am highlighting a paper that uses interesting concepts from different fields to create a cool new tool for cryo-ET. While not exactly using an omics approach, it is a pretty neat use of technology, so I decided to include it here.

In crowded biological environments it is often difficult to identify small features of interest or those without a distinct appearance, especially without prior structural information in cryo-ET workflows. To learn more about cryo-ET, and the history of elcetro microscopy for biological samples see here.

The authors decided to approach this problem by designing a better tag that is suitable to use with living cells, suitable for frozen hydrated samples, does not obscure the biological sample, not rely on stains or metals (hence not scatter more than the atoms of the specimen itself), while providing precise localization with high confidence. And although there are limitations to the applications of their approach, they did come up with nice solution which I think will likely improve with time. What I found really cool about this work was how they combined ideas and tools already developed in other fields, and applied them to design a good (hopefully improved) tool for cryo-ET analysis.

Here is what they do: They first designed a DNA origami tag, and then combined it with RNA aptamers that target a specific protein, that can then be incubated with biological samples.

Never heard of DNA origami? DNA origami is the folding of a long single stranded DNA with multiple small ‘staple’ strands to generate specifically defined 2D and 3D shapes at the nanoscale. See the original paper here. The advantages for using DNA to generate a cryo-ET tag are that Phosphorus (that forms the DNA backbone) scatters elastically ~4x more electrons than carbon, oxygen, or nitrogen, that can provide high enough contrast but not obscure biological structures around it in cellular tomograms. And, DNA origami has been extensively tested for use with cryoEM, so they already knew the DNA tags can handle the Cryo-ET work flow. The DNA origami signpost nanostructure they design is built from a 7,249-nucleotide scaffold strand (single-stranded M13mp18) hybridized to 238 staple oligonucleotides to form a structure of approximately 5 MDa comprising 96 parallel helices arranged in a honeycomb lattice.

Next trick was to target this tag to specific proteins of interest in a biological sample. For this they employed aptamers. Aptamers are RNA oligonucleotides that bind to a specific target with high affinity and specificity, similar to how an antibody binds to an antigen. Read more about aptamers here. The aptamers against common fluorescent proteins such as GFP and mCherry have been previosuly characterized by some wonderful work done here at Cornell in our Provost Mike Kotilikoff’s lab. See that paper here. So the authors decided to test their approach by using the aptamers against GFP/ mCherry to look at proteins that have been tagged with these fluorescent protein tags. In principle, if the workflow can be modified, this could also allow to correlate fluorescence signals with the tomograms.

They then use this tag to test in three applications- to look at proteins in membrane vesicles, proteins on the viral surface, as well as cell surface proteins in mammalian cells. They show that their tag can allow more precise localization of the labeled protein whereas immuno-gold labeling puts the gold clusters up to 20-40 nm away. Their tag was easily identifiable, and in many cases could be directly traced to protein density on the vesicle surface. The latter can be useful for purposes of manual particle picking for sub-volume averaging applications, although getting really detailed structual resolution of labeled proteins will take more work.

That’s really the heart of it! The approach is pretty cool, and does achieve many of the objectives they lay out at the beginning. It isn’t perfect, for instance, it is limited by available specific aptamers, and unclear if it will work for visualization of intracellular organelle proteins rather than just those at the cell surface. But, it’s pretty cool!

Research Paper Highlight: Genetic interaction mapping informs integrative structure determination of protein complexes

Hannes Braberg, Ignacia Echeverria, Stefan Bohn, Peter Cimermancic, Anthony Shiver, Richard Alexander, Jiewei Xu, Michael Shales, Raghuvar Dronamraju, Shuangying Jiang, Gajendradhar Dwivedi, Derek Bogdanoff, Kaitlin K. Chaung, Ruth Hüttenhain, Shuyi Wang, David Mavor, Riccardo Pellarin, Dina Schneidman, Joel S. Bader, James S. Fraser, John Morris, James E. Haber, Brian D. Strahl, Carol A. Gross, Junbiao Dai, Jef D. Boeke, Andrej Sali, Nevan J. Krogan

Science Vol. 370, Issue 6522, (2020); DOI: 10.1126/science.aaz4910

For Part 2 of the Structuromics series I am highlighting a paper that combines high throughout genetic interaction mapping and integrative modeling for in vivo structural profiling.

Large scale genetic interaction (GIs) profiling in model organisms such as budding yeast has been transformative in informing us about cellular pathways and ascribing function to uncharacterized proteins. Applicability of CRISPR-Cas9 based genetic manipulation has made high-throughout genetic interaction profiling possible in mammalian cells as well as almost any model system. In this work, the authors put to test the idea whether genetic interaction profiles can be employed to determine protein structural information.

Traditionally, genetic interaction analysis involves comparing the phenotypes when loss of the protein product of one gene is combined with the loss of another one (for digenic GIs) or multiple (for multigenic GIs). Well, how about generating a library of mutants by mutating every single residue in a protein and combining each with the gene deletion collection? The resulting (hundreds of!) GI profiles or epistatic miniarray profiles (pE-MAPs) can be compared to each other. Residues involved in the same interactions are likely to generate similar GI profiles, which may be distinct with GIs from residues involved in interactions with a different protein. The phenotypic similarities and differences can then be incorporated into integrative modeling algorithms to generate structural models for architectures of protein complexes. Its a tour de force, and the authors show it works!

The quantitative GI profile dataset can be enriched combinatorially, by measuring sensitivity to chemical datasets or by also measuring changes under different stresses. And by employing barcodes and next-gen sequencing approaches the massively-parallel screens can be performed in liquid cultures allowing cost and infrastructure reduction. What is really powerful about this approach is the ability to capture effects of transient interactions that are not otherwise possible to discern from GI datasets. In addition, the structural information infered reflects in vivo protein complex interactions providing a built-in functional readout. There are caveats of course- the choice of amino acid change can have a huge impact on protein stability or protein-protein interactions or that single amino acid changes may illicit only modest effects that may not be captured in GI interactions. The true value of analysis will likely be amplified when combining it with orthogonal approaches- such as cross-linking mass spectrometry or crystallography/ cryo-EM. Extending the analysis to disease associated alleles or for multiple components of protein complexes, too can provide valuable information. Overall, I found it is really amazing how much can be inferred by the approach.

Research Paper Highlight: Dynamic 3D proteomes reveal protein functional alterations at high resolution in situ

Valentina Cappelletti, Thomas Hauser, Ilaria Piazza, Monika Pepelnjak, Liliana Malinovska, Tobias Fuhrer, Yaozong Li, Christian Dorig, Paul Boersema, Ludovic Gillet, Jan Grossbach, Aurelien Dugourd, Julio Saez-Rodriguez, Andreas Beyer, Nicola Zamboni, Amedeo Caflisch, Natalie de Souza, and Paola Picotti.

Cell 184, 545–559, January 21, (2021); DOI: 10.1016/j.cell.2020.12.021

In many ways the tale of blind men and the elephant is a good analogy for our understanding of biological systems. We create tools to probe biomolecules and make interpretations from our observations. But no one experimental design or approach is bias-free. To truly understand function, we need multi-dimensional, holistic analyses from complementary observational viewpoints. Technological improvements have allowed merger of biological disciplines that tradionally used non-overlapping approaches to address the same biological questions. One such case that strikes me as spectacular is what has been termed as ‘structuromics’– a combination of genomic, proteomic, genetic and cell biological approaches to ascertain protein structure and function in the cellular context. There have been multiple really cool papers published in the last few months that I’d like to highlight in a multi-part structuromics series.

Part 1 of the series is a paper that combines proteomics and structural biology by employing limited proteolysis and mass spectrometry to predict large scale, and biologically relevant, structural and conformational changes in proteins in response to stress.

In this work, the authors employ ‘limited proteolysis coupled to mass spectrometry’, or LiP-MS, to obtain information about structural changes in thousands of cellular proteins simultaneously, in response to various stresses. The idea is a simple one - if you subject a protein to proteases, the exposed parts of the protein are readily digested, while the structured/ modified/ complexed part that are hidden are protected. Now, if the protein undergoes structural changes upon binding to a ligand or other proteins, the acessibility and the pattern of protease digestion of the protein will change. The experiment, then, can be repeated under many different conditions, and for many different proteins by employing mass spectrometry to generate quantitative profiles or fingerprints of structural alterations.

Biochemists have used limited proteolysis as a tool to study individual proteins for a long time. What is remarkable about the LiP-MS and other similar approaches is the wealth of information on thousands of cellular proteins that can be obtained at the same time, allowing to see higher order patterns in cellular pathways, which is not possible when studying one protein at a time. Additionally, many proteins aren’t as amenable to pure structural analyses, and these approaches can infer structural information that can guide protein functional studies. By applying their approach to E. coli as well as yeast cells, the authors show that monitoring conformational changes in different conditions provides important functional information that is otherwise missed when looking at protein abundance measurements alone.

The approach is useful on its own, but its richness can be greatly enhanced if it can be combined with complementary workflows. For instance, combining it with other mass spec based readouts- protein abundance, phosphorylation, ubiquitylation, and other modifications, cross-linking based peptide analysis, etc. can tell what mechanisms underlie the conformational changes. Similarly, combining with evolutionary analyses to look at co-evolution rates of amino acid residues within a protein or interacting partners can predict the conservation mechanisms underlying the structural changes. Proteins with low resolution crystal or cryo-EM structures too can be significantly enriched with the knowledge of known conformational changes. There is a lot of room for development of modifications to the approach, and all that can be done is truly exciting!

Research Paper Highlight: Pervasive functional translation of noncanonical human open reading frames

Jin Chen, J. Zachery Cogan, James K. Nuñez, Alexander P. Fields, Britt Adamson, Daniel N. Itzhak, Jason Y. Li, Matthias Mann, Manuel D. Leonetti, Jonathan S. Weissman. Science volume 367, Issue 6482, pp. 1140-1146 (2020); DOI: 10.1126/science.aay0262

From the perspective of gene prediction, it is really important to be able to correctly identify open reading frames (ORFs) that are likely to code for functional proteins. In the simplest of cases the ORF starts with a AUG codon, and ends with a stop codon (UAA/UGA/UAG) and the nucleotide count is divisible by three. Even then, one of the biggest challenges for curation and function annotation, has been to deal with the exceedingly high number of predicted small ORFs (containing less than 100 codons). More than 260,000 small ORFs are predicted in budding yeast alone, and their probability of being biologically meaningless is quite high. At the same time, several really interesting microproteins (encoded by the small ORFs) have been identified and characterized in diverse cellular pathways. The question is that of finding a needle in a haystack- how do we identify functional small ORFs amongst the thousands that are predicted.

One of the biggest strides that helped in this quest of finding these tiny protein coding ORFs, came from the development of a highly sensitive and quantitative tool to assess active ribosomal translation called Ribosome profiling by Nicholas Ingolia and Jonathan Weissman in their 2009 paper. Right away, it became obvious that ribosomes actively translate non-canonical small ORFs, many that don’t even start with conventional start codons- an observation that has been made in every model organism and cell system ever since. But the proof of the actual protein products and their functional significance was still lacking.

In this work, they attempt to address that systematically. They perform ribosome profiling in four different cell lines and detect thousands of overlapping small ORFs with ribosomal footprints. But they go further and find evidence of existence of at least a subset of the protein products by total proteome MS as well as by HLA-I peptidomics. The coolest part of the paper, though, is the preliminary functional analysis they present. By performing pooled CRISPR-KO screens, they provide evidence that loss of many of these small ORF products results in definitive disadvantage to cellular fitness. They follow-up candidates of two categories of non-canoical ORFs: those embedded in annotated lncRNAs (long non-coding RNAs) and small ORFs that exist upstream to a main/ annotated/ canonical ORFs downstream. The most intriguing finding of all is that in most of the cases, the protein products of the small uORFs localize to the same cellular location as the protein product of the main ORF, and show physical association with it. The protein products of lncRNAs also localize to discrete cellular locations suggestive of unique functions.

Intriguing as it is, this work opens the door to allow more detailed characterization of function and regulation of these new microproteins. Do non-AUG start codons in many of these small ORFs provide a unique mechanism for regulating their expression? Are there conserved features in these microproteins that might allow identification of their homologs in other organisms (if they exist)? What is the function of these microproteins? Do the protein products of the small uORFs promote or inhibit the functions of the protein products of the main ORFs? Or if the function of the protein product is truly needed? Could it be another mechanism to regulate the function of the corresponding RNA which encodes it- such as by preventing the lncRNA to perform its regulatory function? Lots of new questions to explore!

Research Paper Highlight: If your P value looks too good to be true, it probably is: Communicating reproducibility and variability in cell biology.

Samuel J. Lord, Katrina B. Velle, R. Dyche Mullins and Lillian K. Fritz-Laylin arXiv:1911.03509v2 (2019)

Later published as a Viewpoint in J Cell Biol (2020) 219 (6): e202001064. DOI:10.1083/jcb.202001064

Cell biologists are no strangers to data variability- growth conditions, age of the cells, and nutrient availability all contribute to sample-to-sample differences. Accurate quantitative analysis and representation of cell biological measurements should ideally be able to reflect the number of times the experiment was repeated, the variability observed within each replicate set for individual measurements, and the key statistical identifiers for each set.

In this preprint, Lord et al, present a convincing case to represent cell biological quantitative data as multi-colored beeswarm or superplots to accurately reflect actual data quality, reproducibility and variability.

Key highlights to note:

1. The problem of comically miniscule (and hence useless) P-values in many cell biology studies is in part due to erroneous and inflated sample size, N. N should represent the number of times an experiment is repeated, not the total number of cells.

2. Bar graphs are problematic because they obscure the distribution of cell-level data as well as the sample-to-sample repeatability.

3. Visual representation of the quantitation should represent variability information about the entire dataset whenever posible. This can be nicely accommodated into color-coded beeswarm SuperPlots. The authors provide detailed ways of making these plots using GraphPad Prism, Excel, R and Python.

For example, if you have three biological replicates of control and treated samples, and you measure the speed of cellular movement of 200 cells in each sample, first calculate the mean of those 200 measurements for each sample, then run a t-test on those sample means (three control, three treated). In the data plot, color-coding the dots by experiment, and plotting the mean as a larger symbol on top of the dots that denote individual measurements, makes it clear that statistical analyses (e.g. error bars and P values) are correctly calculated across separate experiments.

Research Paper Highlight: Excised linear introns regulate growth in yeast.

Morgan et. al., Nature volume 565, pages606–611 (2019)

Introns are generally thought of as a waste by-product of RNA processing and molecules that are rapidly degraded soon after they are produced.They are assumed to be very short lived species- once the spliceosome joins the two exons, the spliceosome disassembles and the intron (lariat structure which looks like a lasso) is released. It is very quickly debranched by Dbr1, polyadenylated by the TRAMP complex and degraded by the nuclear exosome. Most introns don’t accumulate or seem to have any phenotype in log phase cultures in rich media.

The authors perform RNA seq in log phase vs saturated cultures, and show that some of the introns (their best examples are ECM33 and SAC6) are dramatically stablized in saturated cultures. In fact they show by RNA seq and Northern blots that the mature RNA for these is completely non-detectable in saturated cultures but the intron is really stable. They also show that the stability is intrinsic to the intron, and not dependent on the host gene, i.e, moving the intron into a completely different context (in a URA3 reporter), still phenocopies the intron stability phenotype in saturated cultures. Yet, the sequence does not show any conserved motif or RNA structural motif. They observe that the distance between the branch point and the 3’ splice site tends to be shorter for stable introns as opposed to introns that are not stable, and they nicely test that if they extend the length of this distance, the ECM33 intron is no longer stable or if they shorten the length in an otherwise non-stable intron of ACT1, the intron is now stable. They also attempt to understand how these introns are accumulating, and are only slightly successful in understanding it at all. The introns are stabilized after prolonged treatment (4h) with Rapamycin, DTT and tunicamycin but not short treatments (1h). The effect does not depend on a specific nutrient starvation but general secretory stress (not IRE1 dependent, so not UPR), nor mediated by Tor1 effectors, Sch9 or Tap42. So conclusions for these data are, not surprisingly, a bit hand-wavy.

They conclude by showing that accumulation of stable introns is advantageous to cellular fitness in stationary phase, but detrimental during rapid log phase growth. Their model is that TORC1 inhibition results in accumulation of these introns, these introns engage and titrate away the spliceosomal machinery, preventing them from too much splicing (esp ribosomal genes which are particularly enriched in introns), thus reducing the growth rate overall but allowing for better survival in limiting conditions. Overall, I think the observation of stable introns is quite striking, but it opens up so many questions than were attempted to be addressed in the work. For instance, what are the Tor1 effectors involved? Is the splicing machinery really getting titrated away? How are the stable introns specifically protected from degradation? And then, bringing into perspective that only 5% of yeast contain introns, as compared to roughly 95% of human genes. If this is indeed a conserved response, this should have a really important contribution to growth signaling in mammalian cells. But that will be for a later time.

Research Paper Highlight: Translocon Declogger Ste24 Protects against IAPP Oligomer-Induced Proteotoxicity.

Can Kayatekin, Audra Amasino, Giorgio Gaglia, Jason Flannick, Julia M. Bonner, Saranna Fanning, Priyanka Narayan, M. Inmaculada Barrasa, David Pincus, Dirk Landgraf, Justin Nelson, William R. Hesse, Michael Costanzo,AMP T2D-GENES Consortium, Chad L. Myers, Charles Boone, Jose C. Florez, and Susan Lindquist, Cell 173, 1-12, March 2018

Islet amyloid polypeptide (IAPP) is a small peptide secreted by the pancreatic β islet cells. It is thought to regulate the function of the islet cells by inhibiting insulin secretion. It was discovered (and named) due to its tendency to aggregate into insoluble amyloid fibrils in patients with type 2 diabetes (T2D). Subsequent work showed that the cytotoxicity was contributed by smaller, intracellular oligomeric intermediates of IAPP, not the mature amyloid fibrils. Yet what aspect of cellular biology is impacted by these toxic oligomers has remained unclear.

Kayatekin and colleagues report that IAPP-induced proteotoxicity is in part contributed by the clogging of the endoplasmic reticulum (ER) translocon, and can be rescued by translocon associated transmembrane protease Ste24 in yeast or its human homolog ZMPSTE24. Ste24/ZMPSTE24 were recently discovered to directly interact with clogged translocons and cleave jammed proteins.

In order to understand IAPP toxicity, the authors took advantage of the cell death associated with the expression of an oligomeric version of the IAPP peptide in budding yeast, Saccharomyces cerevisiae. Using complementary genetic screens the authors found that functional Ste24 protease was critical for alleviating IAPP toxicity in yeast. The suppression by Ste24 was specific to IAPP and did not mitigate the toxicity caused by other aggregation prone proteins such as α-synuclein or TDP-43. Emphasizing the conservation of the declogging function, overexpression of the human homolog ZMPSTE24 in yeast lacking Ste24 could efficiently rescue IAPP toxicity. Using their yeast model for assaying declogging capability, the authors analyzed 111 ZMPSTE24 single nucleotide polymorphisms discovered in T2D patients and non-diabetic controls and found an enrichment of loss-of-function mutations among people with T2D (although somewhat mild).