Label-free Viability Assay using Holographic Video Microscopy

8 Total Holographic Characterization (THC) is presented here as an efﬁcient, automated, label-free method of accurately identifying cell viability. THC is a single-particle characterization technology that determines the size and index of refraction of individual particles using the Lorenz-Mie theory of light scattering. Although assessment of cell viability is a challenge in many applications, including biologics manufacturing, traditional approaches often include unreliable labeling with dyes and/or time consuming methods of manually counting cells. In this work we measured the viability of Saccharomyces cerevisiae yeast in the presence of various concentrations of isopropanol as a function of time. All THC measurements were performed in the native environment of the sample with no dilution or addition of labels. We compared our results with THC to manual counting of living and dead cells as distinguished with trypan blue dye. Our ﬁndings demonstrate that THC can effectively distinguish living and dead yeast cells by the index of refraction of individual cells.


10
The use of yeast cells, particularly Saccharomyces cerevisiae, is ubiquitous in both industry and academia 1-3 for applications Additionally, with any dye-based staining viability measurements, there is a danger that the dye can interact either with 23 the cells or with another experimental variable in unintended ways. As an example, trypan blue has been shown to adversely 24 interact with cells, often rupturing them and thus rendering viability measurements unreliable 19 . Although the shortcomings of 25 label-based viability assays are well known, few label-free alternatives exist 20 . 26 In this work we introduce a label free, automated technique to reliably measure the viability of yeast using Total Holographic 27 Characterization ® (THC). THC is a technology, based on holographic video microscopy, developed to detect, count, and 28 characterize subvisible particles in suspension 21 . Here we assessed yeast viability using xSight, which implements THC and 29 combines holographic microscopy with microfluidic sample handling to provide precise measurements of particle size and 30 refractive index. We demonstrate that cell refractive index can determine the viability of the cell. Although holographic 31 techniques have been used to study various cell types 22-27 , a fast, automated, label-free viability tool using cell refractive index 32 has not been achieved previously. 33 To demonstrate THC as a new technology for viability studies, we used alcohol to gradually kill yeast cells, as has been done 34 in previous yeast viability studies 28,29 . We show the viability changes of Saccharomyces cerevisiae under various concentrations 35 of isopropanol and compare our results with staining using trypan blue.

37
Holographic approach to yeast viability 38 We employ holographic video microscopy 21, 30-39 to assess yeast viability. In our approach, particles in suspension (such 39 as yeast cells) flow through a microfluidic chip as they are illuminated by a collimated laser beam. Laser light scattered by 40 the particles interferes with the incident laser light, forming an interference pattern called a hologram. A schematic of this 41 technology is shown in Figure 1a. A scatter plot of size on the horizontal axis and refractive index on vertical axis for 4 species of particles: 1.51µm diameter polystyrene spheres (in cyan), 2.56µm diameter polystyrene spheres (in violet), 1.49µm diameter silica spheres (in orange), 2.63µm diameter silica spheres (in yellow). Each point on the plot represents a single particle detected with THC during measurement. The colored boxed are user-defined regions of interest. Particles outside of the 4 user-defined boxes are colored gray.
The holograms are recorded on a camera and fit to Lorenz-Mie theory of light scattering 32, 40-42 . A fast, multi-parameter 43 optimization of the fit then yields various particle parameters including particle size, refractive index, and 3-dimensional 44 position that correspond to that hologram. A key advantage of this approach is that in addition to particle size, THC quantifies 45 the refractive index of each particle, which is indicative of its composition. For example, Figure 1b shows a scatter plot of a 46 sample that consists of a mix of four different particles species: two sizes of polystyrene microspheres and two sizes of silica 47 microspheres. Each point on the plot represents a single particle detected by xSight during measurement. On the horizontal 48 axis is the measured particle diameter and on the vertical axis is the measured particle refractive index. The colored boxes illuminated with a blue laser is 1.61 and a refractive index of 1.43 is consistent with porous silica. It is important to note that 54 with a standard particle sizer, only two populations would be visible in this sample: one at a particle diameter of around 1.5µm 55 and one at a particle diameter of around 2.6µm. Since THC measures refractive index in addition to size, particles of the 56 same size but different composition can be easily distinguished, such as the 1.49µm silica and 1.51µm polystyrene spheres (in 57 orange and cyan respectively) and the 2.63µm silica and 2.56µm polystyrene spheres (in yellow and violet respectively).

58
This technology has similarly been used to distinguish bacteria from plastic beads and oil in water 43 since each of those 59 species is composed of materials of different refractive indexes. A study of porous silica particles showed that THC is able to in water, the refractive index of a given particle will be somewhere between the refractive index of the silica matrix and the 67 refractive index of the water that fills its pores. To what extent the resulting effective refractive index will be closer to silica or 68 closer to water depends on the particle's porosity. Hence it is possible to extract morphological information from the behaviour 69 of the particle refractive index in media of varying refractive indexes. Quantitative morphological information about particle 70 porosity has been studied in porous plastics, protein aggregates, and nanoparticle agglomerates 44, 47 .

71
Yeast cells are also heterogeneous particles with various cell components acting as scattering materials with different 72 refractive indexes. Since cell death often involves physiological changes in cell structure and composition, these changes are   Figure 2. (a) A scatter plot of size on the horizontal axis and refractive index on vertical axis for a yeast sample before the addition of alcohol. Each point on the plot represents a single particle that flowed through the viewing region of the microfluidic chip during THC analysis. The colored boxed are user-defined regions of interest. The points colored in orange represent dead yeast cells and the points colored in cyan represent live yeast cells. Particles outside of the user-defined boxes are colored gray. (b) Density distributions of particle size for the sample shown in (a). The orange, cyan and gray curves represents the size density distributions of dead cells, live cells, and all particles respectively. The area under each curve for a given size range represents the number of particles (of the species represented by that curve) in that size range. The peak of each curve shows the most common size of each particle type. (c) Density distributions of particle refractive index for the sample shown in (a). The coloring is the same as in (b). The area under each curve for a given refractive index range represents the number of particles (of the species represented by that curve) in that refractive index range. The peak of each curve shows the most common refractive index of each particle type. (d)-(f) Scatter plot, size density plot, and refractive index density plot as in (a)-(c) but for a yeast sample that was exposed to 15% isopropanol by volume for 71 minutes.
As can be seen in Figure 2a, while the live and dead cells overlap in size, their refractive indexes are distinct. In this sample 82 live cells were found in the size range of 3.5µm to 6.2µm and refractive indexes between 1.39 and 1.416. Dead cells appeared 83 at slightly smaller sizes: 2.5µm to 5.2µm and significantly larger refractive indexes: between 1.417 and 1.462. The overlap in 84 size and distinction in refractive index of live and dead cells is further illustrated in Figure 2b, and 2c. These plots are size and 85 refractive index density distributions respectively. The size density curve in Figure 2b, for example, is related to the probability 86 for a particle to have a particular size. The area under the curve over any particle size range represents the number of particles 87 3/9 that can be found in that size range. These curves are more informative and more reliable alternatives to histograms. In these graphs, the gray curves represent all particles, the orange curves represent dead cells, and the cyan curves represent live cells.  Comparing viability results using holographic imaging and the TB exclusion assay 103 To explore how yeast respond to different concentrations of isopropanol over time and to compare our holographic approach to 104 the TB exclusion assay, we exposed yeast samples to 0%, 15% and 20% isoproponal by volume, and measured the resulting 105 viability over time using both xSight and trypan blue. Figure 3 shows the results of these studies.   The presence of ambiguously stained cells, as in Figure 4b can also contribute to the above-mentioned higher variability in 128 staining measurements than in measurements with THC.  This conclusion is also consistent with morphological cell changes. If cells shrink due to expulsion of low index of refraction 135 water and the remaining cell sub-components have a higher index of refraction than water, then the net index of refraction of the 136 cell will increase as observed in our measurements. THC characterizes each particle that flows through the xSight microfluidic 137 chip and analyzes it as a sphere. The multi-parameter optimization finds an effective sphere with a hologram most similar to the 138 hologram of the particle being analyzed 44, 46 . A smaller cell with stronger scattering components will be identified as a smaller  Our holographic approach is also a promising way to study the progression towards cell death or other stress responses.

162
Yeast preparation 163 The yeast solution was made using glucose (SIGMA brand, item G8270-1KG), filtered deionized (DI) water, and dry instant 164 yeast (Amish Market store brand, New York, NY). A glucose stock mixture was made by mixing 25g of DI water with 0.5g of 165 glucose until the glucose was fully dissolved. Once dissolved, the glucose mixture was filtered with a 60cc luer lock disposable 166 0.45µm syringe filter (Millex brand). If not made for immediate use, the glucose mixture was stored in a 4 • C refrigerator.

167
The glucose solution was diluted with filtered DI water in a 1:1 proportion, using 4mL of each, in a 30mL vial. Using