Dextranol: A better lyoprotectant

Dextranol, a reduced dextran, prevents damage to stored dry protein samples that unmodified dextran would otherwise cause. Lyoprotectants like the polysaccharide dextran are critical for preserving dried protein samples by forming rigid a glass that protects entrapped protein molecules. Stably dried proteins are important for maintaining critical information in clinical samples like blood serum. However, we found that dextran reacts with serum proteins during storage, producing high-molecular weight Amadori-product conjugates. These conjugates appeared in a matter of days or weeks when stored at elevated temperatures (37° or 45°C), but also appeared on a timescale of months when stored at room temperature. We synthesized a less reactive dextranol by reducing dextran’s anomeric carbon from an aldehyde to an alcohol. Serum samples dried in a dextranol- based matrix protected the serum proteins from forming high-molecular weight conjugates. The levels of four cancer-related serum biomarkers (prostate specific antigen, neuropilin-1, osteopontin, and metalloproteinase 7) decreased, as measured by immunoassay, when serum samples were stored for one to two weeks in dextran- based matrix. Switching to a dextran-based lyoprotection matrix slightly reduced the damage to osteopontin and completely stopped any detectable damage during storage in the other three biomarkers when for a period of two weeks at 45°C. Dextranol offers a small and easy modification to dextran that significantly improves the molecule’s function as a lyoprotectant by eliminating the potential for damaging protein-polysacharide conjugation.


36
Room temperature protein preservation is important for food, biologics, purified protein products (e.g.

79
In our work to preserve clinical serum samples by isothermal vitrification, dextran is essential since it 80 gives the dried samples the high glass transition temperature needed to maintain stability at room temperature.

81
However, conjugation of biomarker proteins with lyoprotectants like dextran could be very detrimental to 82 downstream proteinaceous biomarkers analysis, especially when analytical techniques that are based on specific 83 binding (such as ELISA) are to be utilized to detect activity. In this work, we analyzed the damage to vitrified 84 human serum proteins caused by conjugation with dextran during dried state storage, and we found that 85 replacing dextran in our lyoprotectant formulation with a reduced dextran (dextranol) (Fig 1)

139
We electrospun the solution into microfibers across a voltage differential in a controlled environment.

140
We filled 1mL syringes with the lyoprotectant cocktail and affixed a stainless steel 18-gage 0.

203
We found that when we used dextran-based lyoprotectant matrix to preserve blood serum by isothermal 204 vitrification, serum began to show signs of damage over time. The damage was detected by the appearance of 205 high-molecular weight smearing seen in gel electrophoresis. While freshly vitrified (and immediately 206 reconstituted) serum appears identical to fresh or frozen serum on gel electrophoresis, after sixteen weeks the 207 smearing becomes very prominent (Fig 2), especially when the samples are stored at elevated temperatures.

208
Individual protein bands are less detectable due to decreased intensity and smearing. While smearing is possibly 209 worse in native protein gels (Fig 2A), denatured (Fig 2B), and denatured & reduced (Fig 2C)

225
High molecular weight smearing and glycosylation increased both with storage time and storage 226 temperature and were accompanied by increased solubility. Smearing was faint after only 1 month of storage at 227 room temperature, but increased when either storage temperature was increased to 37°C or storage time was 228 extended to six months (Fig 3A). These effects were additive as after six months of storage at elevated 229 temperature the smearing and decrease in individual protein band contrasts was significantly worse than storage 230 either for six months at room temperature or for one month at elevated temperature. Concurrent with the 232 indicates that high-molecular weight bodies in the smear are becoming significantly glycosylated. We also 233 noticed that samples that had become more soluble and would not precipitate efficiently using a standard TCA 234 precipitation protocol (S1 Fig, S1 Table).  (Figs 4 and S4). Samples that were frozen, freshly vitrified in dextran-based matrix or dextranol-247 based matrix were all virtually indistinguishable from fresh, never frozen serum when analyzed by gel 248 electrophoresis under either native (Fig 4A) or denaturing conditions (Fig 4C). However, after 140 days of 249 storage at 37°C, vitrified samples in dextran-based matrix showed significant smearing, while dextranol-based 250 matrix samples were almost indistinguishable from frozen or fresh serum. Dextranol also protected isothermally vitrified serum and BSA when stored at high temperatures (45°C).

260
The glass transition temperature of the vitrified serum in either dextran or dextranol-based matrix was 50-55°C to a frozen control, but after storage for 1-2 weeks at 45°C, the band from the main BSA monomer was highly 267 diminished, largely replaced by a high-molecular weight smear (Fig 5A). Dextranol-based matrix was able to 268 effectively eliminate this damage, with the sample stored at 45°C for two weeks looking indistinguishable from 269 the freshly vitrified sample and very similar to the frozen aliquot. 278 In addition to protecting BSA, the dextranol-based matrix also protected human serum proteins when 280 isothermally vitrified and stored at higher temperature. Most protein bands became largely diminished when 281 the sample was stored for one or two weeks in a dextran-based matrix (Fig 5B), and were replaced by a high-282 molecular weight smear, but storage in dextranol-based matrix protected the proteins. The serum proteins look 283 indistinguishable from freshly vitrified serum after 2 weeks at 45°C, and looked very similar to frozen serum.

284
In addition to providing overall protection to high abundance serum proteins, and albumin specifically, 285 we also found that dextranol-based lyoprotectant matrix provided much better protection than the dextran-286 based matrix for four individual proteinaceous biomarkers in the high-temperature stored serum samples. We 287 selected 4 cancer biomarkers to be tested for stability. All four biomarkers that were examined (prostate specific 288 antigen, neuropilin-1, osteopontin, and metalloproteinase 7) showed losses (as measured by ELISA) after one 289 or two weeks of storage at 45°C (Fig 6 and S2 Table). In dextranol-based matrix, PSA (Fig 6A) and neuropilin 290 (Fig 6B) levels on day 1 were slightly reduced, potentially due to the drying process (by 7 and 8%, respectively), 291 but were stable during seven or fourteen days of storage. MMP-7 levels (Fig 6D) in dextranol-based matrices formulations containing dextrans cause increase in protein size, increases solubility, and altered acidity, but this 319 was interpreted as aggregation or damage, and not attributed to dextran-protein conjugation. Studies using size 320 exclusion chromatography observe increased protein size after storage with dextran and speculate that this is 321 caused by "soluble aggregation" (50-56). However, size exclusion chromatography cannot distinguish dextran 322 conjugated proteins from dimeric or oligomeric protein "aggregates". Thus, the "soluble aggregates" reported 323 may instead be dextran conjugates (potentially, in addition to protein self-aggregation