Lab-on-a-disc Screening

The results we have described so far have focused on developing libraries of nonribosomal peptides (NRP), but we have not yet described how these libraries may be screened. High throughput screening is essential to couple MAGE to improved drug design. In the laboratory, we have worked with tyrocidine and surfactins. They are NRP antibiotics, which are cytotoxic to red blood cells [1][2][3]. In our first generation screening experiments, we tested if we could use this property to screen NRP libraries for reduced cytotoxicity.

Achievements

• UV-Vis absorbance spectra proved that red blood cells are lysed in the presence of certain NRPs. However, the experiments should be repeated with fresh blood cells. This means lab-on-disk experiments are feasible for screening mutated NRPs.

Background

Certain groups of nonribosomal peptides, including the commonly used antibiotic tyrocidine as well as surfactin, display non-specific cytotoxicity [1][2][3]. For this reason, these antibiotics are commonly used topically, as intravascular injection causes hemolysis of red blood cells [1]. Low concentrations of NRP antibiotics have been found to be fatal to animal models, which further supports the need for an in-vitro screening. Through UV-Vis spectrophotometry, we are able to analyze the amount of light absorbed in a solution with the addition of our compounds of interest. The absorbance spectra of red blood cells from 300-800 nm changes depending on the state of the red blood cells [4][5].

Methods

To study cell lysis, the absorbance of fresh blood cells was measured without NRP products in comparison to fresh blood spiked with tyrocidine. UV-Vis absorption spectra were recorded between 300-800 nm using three different devices; in cuvettes, by nanodrop, and on a DS-11 + Spectrophotometer (DeNovix). The concentration of NRP was varied between samples to determine cytotoxicity of tyrocidine and sensitivity of red blood cells. A detailed description of the experiments can be found under Lab NoteBook.

Results

The UV-Vis absorption spectra of red blood cells suspended in the absence or presence of antibiotic (Figure 1) proves toxicity of tyrocidine towards blood cells. For our test the antibiotic used was Tyrothricin, which is composed of 5 different tyrocidine varieties and gramacidin produced by Brevibacillus parabrevis. The drop of the green curve around 600 nm indicates that red blood cells absorbs less at this wavelength. When the curve reaches 0,  it is an indication of cell lysis.
 

Figure 1. Comparison of blood cells in different buffers.
grey line: blood cells anticoagulated in heparin with 0.15M NaCl; non-lysed
green line: blood cells anticoagulated in heparin with unknown concentration of Tyrothrycin; lysed

Decrease of UV absorbance around 600nm occurred in all of the blood samples with tyrothricin indicating lysis of the sample (Figure 2).

Figure 2. Comparison of blood cells in presence of antibiotics over time.  
pink and black line : unknown concentration of Tyrothrycin measured after 30sec with blood cells anticoagulated in heparin; green line: unknown concentration of Tyrothrycin measured after 120sec with blood cells anticoagulated in heparin.

Nanodrop measurements of red blood cells spiked with tyrothricin at different concentration:

Figure 3. Comparison of blood cells (anticoagulated with heparin) in presence of 0,15M NaCl and different concentrations of tyrothricin after 45-60sec of incubation. Nanodrop measurement.

In all samples, incomplete cell lysis was observed (Figure 3). The lysis may be incomplete due to the shorten incubation time of 45-60 seconds. The blood sample resuspended in 0.15M NaCl also shows turbidity at 600-700nm, despite it being an isotonic solution. The partly cell lysis may be explained by either the pipette tip bursting the cells (mechanical forces) or due to hemolysis of the cells. The blood cells were at the point tested three days old, which may lead to bursting[5]. We can conclude that the antibiotic lyses the red blood cells, but we cannot determine its specificity very accurately.

It was also possible to test lysis of red blood cells using an extract from Brevibacillus parabrevis (Figure 4). Prescence of tyrocidine was confirmed by LC/MS and MALDI-TOF prior to the blood cell analysis. It is of interest to see, if cell extracts alone can be used to screen for cytotoxicity as it is easy to prepare cell extractions.

Figure 4. UV-vis absorption spectra of blood anticoagulated with heparin in presence of unknown concentrations of tyrocidine after 45-60sec of incubation. Nanodrop measurement.

In this experiment, cell lysis is also observed in all samples as with the standard. The control sample should retain blood cell stability, yet it did not. This may be due to the same factors as previously discussed. The data shows that further experiments are required to confirm cell lysis across samples. The high peaks of a few samples might suggest complete cell lysis is due to the presence of antibiotics.

Discussion

The results indicates, despite their uncertainity, that screening of on blood samples can be used as method for screening. Litterature supports lab-on-a-disc screening as having strong potential to be a low-cost, quick, and simple way of detect and screening libraries of NRP antibiotics. For this reason, we have described the basis of a lab-on-disc device.

Concept of lab-on-a-disc

The rotating disc would be an easy and low-cost way to screen bacterial colonies. Our inspiration came from the start-up company BluSense Diagnostics, who specialise within microfluidics and rotor devices. The rotor disc would enable visual screening (detection) of bacterial colonies by mixing both small volumes of red blood cells and supernatants of bacterial culture. The primary idea of the disc would enable us to screen five different colonies at the same time. Figure 5., below, displays a simple set-up of lab-on-a-disc device.

Figure 5. Illustration of lab-on-a-chip concept for screening the MAGE products.

The use of the disc for observing cell lysis is very easy (Figure 5.). According to methods used in BluSense Diagnostics, it consists of a few steps. First, loading the bacterial supernatant and blood volume to specific inlets (Figure 6). When the disc starts to rotate, the bacterial supernatant is moved to the mixing/detection chamber where a certain volume of the blood sample is measured out and then moved (with the next rotation) to a mixing chamber, where it is mixed with bacterial supernatant (Figure 7.A). Subsequently, the disc stops rotating, before again beginning to accelerate and decelerate for a certain time period to mix the fluid (Figure 7.B). The disc stops and after a certain time period, the blood cells can be analysed by UV-Vis as described earlier to determine, if the sample contains cytotoxic antibiotics.

Figure 6.  Visual presentation of an original disc design. Disc provided by BluSense Diagnostics.

Figure 7.  Visualisation of rotations influencing  samples’ microflow.
A)Two different solutions are transferred to the mixing chamber. B) Mixture after several rotations of the disc. C) Visible cell lysis. 
Pictures provided by BluSense Diagnostics.

Further investigation is still required to determine: sensitivity the test, the interaction between bacterial supernatant-red blood cells, the most efficient method of extraction of NRPs for screening, and optimization parameters for mixing and microflow within the disc. However, it is very likely that this design would be an applicable screening method for MAGE. 

References

1. Dubos, R. J. (1941). THE PRODUCTION OF BACTERICIDAL SUBSTANCES BY AEROBIC SPORULATING BACILLI. Journal of Experimental Medicine, 73(5), 629–640. doi:10.1084/jem.73.5.629
2. Findlay R. D., Taeusch H. W., David-Cu R., Walther F. J., Pediatr Res., Lysis of red blood cells and alveolar epithelial toxicity by therapeutic pulmonary surfactants.(1995); 37(1):26-30.
3. Pape W. J., Pfannenbecker U., Hoppe U. (1987-1988 Fall) Validation of the red blood cell test system as in vitro assay for the rapid screening of irritation potential of surfactants.Mol Toxicol.;1(4):525-36
4. Jiang, N., Tan, N. S., Ho, B., & Ding, J. L. (2007). Measurement of the red blood cell lysis by bacterial hemolysin. Protocol Exchange. doi:10.1038/nprot.2007.483
5. Jávorfi T., Erostyák J., Gál J., Buzády A., Menczel L., Garab G., Razi Naqvi K. J. Photochem Photobiol B. (2006) Quantitative spectrophotometry using integrating cavities. 82(2):127-31. Epub 2005 Nov