In vitro Measurement of CMP-Sialic Acid Transporter Activity in Reconstituted Proteoliposomes.

Nucleotide-sugar transporters (NSTs) facilitate eukaryotic cellular glycosylation by transporting nucleotide-sugar conjugates into the Golgi lumen and endoplasmic reticulum for use by glycosyltransferases, while also transferring nucleotide monophosphate byproducts to the cytoplasm. Mutations in this family of proteins can cause a number of significant cellular pathologies, and wild type members can act as virulence factors for many parasites and fungi. Here, we describe an in vitro assay to measure the transport activity of the CMP-sialic acid transporter (CST), one of seven NSTs found in mammals. While in vitro transport assays have been previously described for CST, these studies failed to account for the fact that 1) commercially available stocks of CMP-sialic acid (CMP-Sia) are composed of ~10% of the higher-affinity CMP and 2) CMP-Sia is hydrolyzed into CMP and sialic acid in aqueous solutions. Herein we describe a method for treating CMP-Sia with a nonselective phosphatase, Antarctic phosphatase, to convert all free CMP to cytidine. This allows us to accurately measure substrate affinities and transport kinetics for purified CST reconstituted into proteoliposomes.

www.bio-protocol.org/e3551 Bio-protocol 10(06): e3551. DOI: 10.21769/BioProtoc.3551 Additionally, because NSTs are virulence factors for pathogens, they are potential targets for antiparasitic and antifungal drugs (Descoteaux et al., 1995;Ma et al., 1997;Hong et al., 2000;Engel et al., 2009;Caffaro et al., 2013;Liu et al., 2013). Studies have also shown that blocking NSTs can inhibit tumor metastasis, as altered cell surface protein glycosylation profiles are often a feature of cancerous cells (Caffaro and Hirschberg, 2006 Given the importance of NSTs, it is necessary to not only understand how they transport their physiological substrates, but also how mutations and potential inhibitors affect their transport activity. Functional characterization of these transporters via transport assays and other means is essential in understanding genetic pathologies and is a key component in drug development. Herein we describe an in vitro method for measuring the uptake of CMP-sialic acid (CMP-Sia) into proteoliposomes reconstituted with the CMP-Sia transporter (CST), one of seven known NSTs found in humans. While developing this method, we realized that commercial stocks of CMP-Sia contain approximately 10% CMP (Ahuja and Whorton, 2019). This observation, coupled with the known fact that CMP-Sia is hydrolyzed in aqueous solution to CMP and sialic acid (Beau et al., 1984;Horenstein and Bruner, 1996), presented a problem for structural and functional characterization of CST because the affinity of CMP towards CST is approximately 100 times higher than that of CMP-Sia (Ahuja and Whorton, 2019). For transport assays, this abundance of CMP would lead to errors in determining the affinity and transport kinetics of CMP-Sia-an issue that, to our knowledge, had not been addressed in the literature. While methods have been described to purify CMP from CMP-Sia (Beau et al., 1984), we ultimately decided to not pursue these since significant amounts of CMP would still be generated through CMP-Sia hydrolysis during long-duration experiments (e.g., multi-day crystallization trials). In addition, although several CMP-Sia derivatives have been described which are resistant to hydrolysis (Burkart et al., 2000;Kajihara et al., 2011;Watts and Withers, 2004), we decided not to employ these since they may have different affinities and transport kinetics than unmodified CMP-Sia, and because they are not commercially available and would thus require custom synthesis. We therefore developed a method of using a nonselective nucleotide phosphatase, Antarctic phosphatase (AnP), to convert all CMP in CMP-Sia solutions to cytidine, which does not have a measurable affinity for CST and would not affect the outcome of functional characterization studies. This approach has allowed us to determine reliable affinity and transport rate constants for CMP-Sia transport by CST. We anticipate that this approach would also be useful for studies of other aspects of glycosylation machinery that require CMP-free solutions of CMP-Sia. Although CMP-Sia is the only nucleotide sugar known to rapidly hydrolyze in aqueous solutions, since AnP is nonselective, it may also be useful for the study of other NSTs if commercial stocks of their nucleotide-sugar substrates also contain high levels of a higher-affinity NMP.    d. We determined that the rate of hydrolysis of CMP-Sia in the transport assay buffer (Buffer C) is 0.005%/min at room temperature. This should be a factor to consider when designing and interpreting data from transport assay experiments. However, given that most of our transport experiments used relatively low concentrations of CMP-Sia and short incubation times, we did not consider the small amount of CMP generated to be significant.

H]CMP transport, and free Sia has a non-detectable affinity for CST and is most likely not transported; 2) the [ 3 H]CMP-Sia is only included as a tracer and is typically diluted at least 100x
with cold CMP-Sia for the assay; therefore, the relative concentration of any CMP will be very low.
5. Incubate the mixture for the desired time and temperature depending on the experiment. We typically incubate at 23 °C for convenience and to minimize CMP-Sia hydrolysis. At this temperature, the vesicles appear to reach equilibrium by 10 min, with the rate of uptake being linear up until 30 s. , it helps to first pre-wash the filters with 2 ml ice-cold Buffer C before applying the vesicles. After the vesicles have filtered through, wash the filters three times with 2 ml of ice-cold Buffer C. We typically use a 12-well filter vacuum manifold (MilliporeSigma) that accommodates 25 mm diameter filters.
8. Place filters in 7 ml scintillation vials, add 5 ml scintillation cocktail, and count in a liquid scintillation counter.

Data analysis
1. To determine background counts, use values from protein-containing vesicles mixed with the hot substrate and then immediately quenched by adding 0.6 ml ice-cold Buffer C. Alternatively, protein-free vesicles that undergo the same experimental conditions as protein-containing vesicles can be used. We have found that these approaches yield similar results.
2. Subtract the background counts from the total counts for each sample to determine specific counts and convert to mol/min for substrate transport.
3. Fit data to a Michaelis-Menten model to determine Km and Vmax (see Figure 1 for example data).