The plant Golgi apparatus modifies and sorts incoming proteins from the endoplasmic reticulum (ER) and synthesizes cell wall matrix material. Plant cells possess numerous motile Golgi bodies, which are connected to the ER by yet to be identified tethering factors. Previous studies indicated a role for cis-Golgi plant golgins, which are long coiled-coil domain proteins anchored to Golgi membranes, in Golgi biogenesis. Here we show a tethering role for the golgin AtCASP at the ER-Golgi interface. Using live-cell imaging, Golgi body dynamics were compared in Arabidopsis thaliana leaf epidermal cells expressing fluorescently tagged AtCASP, a truncated AtCASP-ΔCC lacking the coiled-coil domains, and the Golgi marker STtmd. Golgi body speed and displacement were significantly reduced in AtCASP-ΔCC lines. Using a dual-colour optical trapping system and a TIRF-tweezer system, individual Golgi bodies were captured in planta. Golgi bodies in AtCASP-ΔCC lines were easier to trap and the ER-Golgi connection was more easily disrupted. Occasionally, the ER tubule followed a trapped Golgi body with a gap, indicating the presence of other tethering factors. Our work confirms that the intimate ER-Golgi association can be disrupted or weakened by expression of truncated AtCASP-ΔCC and suggests that this connection is most likely maintained by a golgin-mediated tethering complex.
Arabidopsis, endomembrane system, endoplasmic reticulum, golgin, Golgi apparatus, optical tweezers, secretory pathway, tethering factor
The architecture of the Golgi apparatus is distinct and seemingly simple. It is an organelle composed of lipids and proteins, arranged as a polarized stack of flattened cisternae, capable of processing and distributing secretory cargo around and out of the cell (Staehelin and Moore, 1995; Polishchuk and Mironov, 2004; Klumperman, 2011). However, the exact mechanisms of Golgi stack assembly and maintenance are still not fully understood (Wang and Seemann, 2011). It is clear that these processes depend on a highly complex and tightly regulated cascade of molecular events (Altan-Bonnet et al., 2004), in which proteins attach to correct membranes and precisely orchestrate a multitude of events including tethering, fusing and budding (Hawes et al., 2010; Wilson and Ragnini-Wilson, 2010).
A Golgi stack has a cis-face through which it receives secretory cargo proteins from the endoplasmic reticulum (ER, Robinson et al., 2007; Hawes et al., 2008; Lorente-Rodriguez and Barlowe, 2011; Robinson et al., 2015) and a trans-face where protein cargo exits via the trans-Golgi network and enters intracellular or exocytotic post-Golgi transport routes (Foresti and Denecke, 2008; Park and Jurgens, 2011). Secretory cargo proteins move through the stack to be processed sequentially and glycosylated by resident N-glycosyltransferases (Schoberer and Strasser, 2011). COPII-coated membrane carriers function in anterograde ER-to-Golgi transport, whereas COPI-coated vesicles transport proteins backwards within the stack and from the cis-Golgi stack back to the ER for recycling (Robinson et al., 2015).
Golgi structure differs significantly between kingdoms. The mammalian Golgi apparatus is most often organized as a stationary peri-nuclear ‘Golgi ribbon’ in which single stacks appear to laterally fuse to create a ribbon-like structure (Nakamura et al., 2012). Plant cells on the other hand contain numerous discrete and highly mobile Golgi bodies (Hawes and Satiat-Jeunemaitre, 2005), which move along the actin cytoskeleton (Boevink et al., 1998; Nebenführ et al., 1999) in a myosin-dependent manner (Sparkes, 2010).
In leaf epidermal cells, Golgi bodies appear intimately associated with ER exit sites (ERES), which are specialized subdomains of the ER at which protein export occurs. This association has resulted in the adoption of the ‘mobile secretory unit concept’ (daSilva et al., 2004; Hanton et al., 2009; Robinson et al., 2015), where the ERES and the Golgi bodies move as a functional unit either with the surface of the ER or over the ER. A study using optical tweezers in living leaf epidermal cells confirmed this concept by demonstrating a strong physical connection between ER tubules and Golgi bodies upon micromanipulation of the latter (Sparkes et al., 2009b).
However, to date we have no definite information on the nature of the molecular complexes that are assumed to be involved in tethering Golgi stacks to ERES. In mammalian cells the golgins, a family of Golgi-localized proteins with long coiled-coil domains, participate in tethering events at the Golgi (Barr and Short, 2003; Short et al., 2005; Barinaga-Rementeria Ramirez and Lowe, 2009; Wong and Munro, 2014). Their coiled-coil domains form a rod-like structure that protrudes into the cytoplasm and thus is free to interact with membranous structures, such as cargo carriers and neighbouring cisternae, or form a part of larger protein tethering complexes (Gillingham and Munro, 2003; Malsam and Söllner, 2011; Chia and Gleeson, 2014). Indeed, Glick (2014) has suggested a new model for ERES/Golgi interaction, where protein tethers could link nascent COPII vesicles budding from the ERES to cis-Golgi membranes or the intermediate compartment that will be transported to the cis-Golgi.
Plants possess a set of putative golgins that locate to Golgi bodies and protein interaction partners have been identified for some of them (Gilson et al., 2004; Latijnhouwers et al., 2005b; Renna et al., 2005; Latijnhouwers et al., 2007; Matheson et al., 2007; Osterrieder, 2012). Their subcellular functions largely remain unclear, although a mammalian p115 homologue has been suggested to be a tethering factor involved in anterograde transport from the ER (Takahashi et al., 2010). A cis-Golgi localized golgin and a good candidate protein for tethering Golgi bodies to ER exit sites is AtCASP (Latijnhouwers et al., 2005a; Renna et al., 2005; Latijnhouwers et al., 2007). AtCASP is a type II transmembrane domain protein with a topology similar to the animal CASP protein (Gillingham et al., 2002). Its N-terminal coiled-coil domains are predicted to form a rod-like structure reaching into the cytoplasm, while its C-terminus contains a transmembrane domain sufficient for Golgi targeting (Renna et al., 2005) and multiple di-acidic DXE motifs required for ER export (Hanton et al., 2005).
CASP, initially identified as a nuclear alternative splicing product of CUTL1 that encodes the transcriptional repressor CCAAT displacement protein CDP/cut (Lievens et al., 1997), was found to locate to Golgi membranes by Gillingham and colleagues (2002). The authors observed protein interactions between CASP and golgin-84 and hSec23 at substochiometric levels, as well as genetic interactions between the yeast CASP homologue COY1 and the SNAREs Gos1p and Sec22p, suggesting a role for CASP in membrane trafficking. Subsequently, Malsam and colleagues reported CASP to function in an asymmetric tethering complex with Golgin-84, with CASP decorating Golgi membranes and Golgin-84 COPI vesicles (Malsam et al., 2005).
Our previous studies indicated a role for AtCASP in Golgi stack formation at an early stage and possibly at the level of ERES (Osterrieder et al., 2010). After Golgi membrane disruption using an inducible GTP-locked version of the small COPII GTPase SAR1, GFP-AtCASP co-localized with Sar1-GTP-YFP in punctate structures on the ER (Osterrieder et al., 2010). AtCASP also labelled reforming Golgi bodies, before Golgi membrane markers, after washout of the secretory inhibitor Brefeldin A (Schoberer et al., 2010).
In this study we used full-length and coiled-coil deletion mutant versions of AtCASP in conjunction with laser tweezers (Sparkes, 2016) to assess its potential role in ER-Golgi tethering and protein transport. Our findings implicate a role for AtCASP in tethering at the ER-Golgi interface, as overexpression of a dominant-negative truncation interferes with the stability of the ER-Golgi connection. However, our observations also suggest the involvement of additional and as yet uncharacterized tethering factors.
Materials and Methods
Standard molecular techniques were used as described in Sambrook and Russel (2001). Fluorescent mRFP fusions of full-length AtCASP and truncated AtCASP-ΔCC were created using the previously published pENTR1A clones (Latijnhouwers et al., 2007) using Gateway® cloning technology according to the manufacturer’s instructions (Life Technologies). Contructs were cloned into the binary expression vector pB7WGR2 (Karimi et al., 2002). Constructs were sequenced and transformed into the Agrobacterium tumefaciens strain GV3101::mp90.
Transient expression of fluorescent protein fusions in tobacco plants
Transient expression of fluorescent protein fusions in tobacco leaves was carried out using Agrobacterium-mediated infiltration of Nicotiana tabacum sp. lower leaf epidermal cells (Sparkes et al., 2006). Plants were grown in the greenhouse at 21°C and were used for Agrobacterium infiltration at the age of 5–6 weeks. Leaf samples were analysed 2–4 days after infiltration.
Stable expression of Arabidopsis thaliana plants
Stable Arabidopsis thaliana plants were created using the Agrobacterium-mediated floral dip method (Clough and Bent, 1998). Arabidopsis plants from a stable GFP-HDEL line (Zheng et al., 2004) were transformed either with mRFP-AtCASP or mRFP-AtCASP-ΔCC and grown on solid ½ Murashige and Skoog medium with BASTA selection. All experiments were performed in T3 or T4 seedlings. As a control, the previously described Arabidopsis line expressing the Golgi marker STtmd-mRFP and the ER marker GFP-HDEL was used (Sparkes et al., 2009b).
Confocal laser scanning microscopy
High-resolution confocal images were obtained using an inverted Zeiss LSM510 Meta confocal laser scanning microscope (CLSM) microscope and a 40x, 63x or 100x oil immersion objective. For imaging GFP in combination with mRFP, an Argon ion laser at 488 nm and a HeNe ion laser at 543 nm were used with line switching, using the multitrack facility of the CLSM. Fluorescence was detected using a 488/543 dichroic beam splitter, a 505–530 band pass filter for GFP and a 560–615 band pass filter for mRFP.
Optical trapping was carried out in stable Arabidopsis lines, using 1) a commercially available dual colour system at Wageningen University, The Netherlands, comprising a 1063nm, 3000mW Nd:YAG laser (Spectra Physics) and x-y galvo scanner (MMI, Glattbrugg, Switzerland) attached to a Zeiss Axiovert 200M with a Zeiss LSM510 Meta confocal laser scanning system (Sparkes et al., 2009b), and 2) a custom-built TIRF-Tweezer system at the Central Laser Facility, Harwell (Gao et al., 2016).
Golgi bodies were trapped using a 1090 nm infrared laser with intensity between 50 and 130 mW. For the ‘100 Golgi test’, Golgi bodies were scored as trapped if the laser beam could move them.
Latrunculin B treatment
To inhibit actin-myosin based Golgi movement, which was required during optical trapping with the confocal microscope trap set up, Arabidopsis cotyledonary leaves were treated with the actin-depolymerising agent 2.5 μM latrunculin B for 30 min as previously described (Sparkes et al., 2009b). Optical trapping experiments were performed within a time scale of 2 h after latrunculin B application.
Tracking and statistical analysis of Golgi body and ER dynamics
Videos for analysis of Golgi body dynamics in stable Arabidopsis lines were taken with a 63x PlanApo 1.4 NA oil objective at 512 × 512 resolution, optical zoom of 3.7 over a region of interest sized 244 × 242 pixels, and recorded for 50 frames at 0.9 frames/sec. Individual Golgi bodies were tracked using ImageJ processing package Fiji (Schindelin, et al., 2012) and the tracking plugin MTrackJ (Meijering, et al., 2012
My flippant response to this question is that it is not plagiarism because, by definition, plagiarism is stealing the words, work, or ideas of another. If the thesis and publications are entirely your own work, and claimed solely by you, then self-plagiarism is an oxymoron.
The question of whether it is “duplicate publication”, however, is one that is not easy to answer definitively. That’s clear by the fact that, although this poll shows that the majority us think it’s ok to publish the thesis text verbatim in an academic journal, one-third of us don’t.
It’s worth thinking about why duplicate publication is frowned upon, and how this impacts theses. Duplicate publication in journals—one person publishing the same words/data more than once in multiple papers—is wrong because it is seen as double-dipping. We all accept the idea that papers should contain unique contributions and that once a paper is published, identical contributions by that individual are superfluous, and merely a method to increase apparent scientific output without much of an increase in effort. And back in the day when journals held the copyright on your words/data, the process was especially problematic, because if you published your own words twice (without the journal granting you license to use them) you would be acting illegally, infringing on the journal’s copyright and arguably carrying out an act of plagiarism.
With regards to theses, the rules of engagement become a bit murkier. I learnt this first hand a number of years ago when I was involved in a new graduate program. We were very keen to embrace the power of the internet at the time, and were initially excited about the prospect of theses appearing on line as a resource. In the end, however, we decided against it. We consulted journals and were universally told that they were not interested in publishing work that had appeared in any public forum before, including online. They were concerned about copyright, and the higher profile rags were also concerned that the appearance of work in a thesis stripped it of its novelty, a practice that would ultimately erode their place as “the place” to access all the hot new science. The former concern has been largely offset by the change in how journals deal with copyright, but the latter concern is real and perhaps all the more important in the days when the stakes are very high and the internet is our main source of scientific exchange.
I voted “yes” here because I believe theses should be an exception to any definition of duplication publication. For one thing, many students don’t have the chance opt out of online thesis publication. We do not offer that privilege to students at my current institution (although fortunately bureaucratic wheels turn slowly and are particularly susceptible to passive aggressive students who want to protect their intellectual property). I do not believe is is fair to penalize a student that may not be able to prevent publication of their thesis, or may not have the forethought to block this practice. I also point out that—unlike publication in a journal—a scientist receives no “career points” for having their thesis appear online, so it clearly not an example of unscrupulous double-dipping. Copyright should no longer be a problem. And although exposing your best data to the world via your online thesis can certainly leave you open to intellectual theft, my experience is that most theses are hard to find online and thus their appearance on the web does not significantly usurp their novelty.
The best use of my thesis is that it levels the sofa at my dad’s house, and I do tend to pine for the days when theses were private documents and not for broad public dissemination. But here we are. So my advice is to publish your thesis in a journal if you can—but please do it only once.