PRECISION OF DISPENSING USING V&P WANDS AND MANIFOLDS

White Paper by Duong Chau V&P Scientific Staff Scientist 

Abstract

V&P Scientific wands and manifolds are designed for the purpose of aspirating from microplates, washing microplates, and dispensing into microplates.  This experiment was set up to determine the precision of the latter capabilities by measuring the quantity of aqueous FITC dispensed per well and calculating the coefficient of variations among the wells as well as among the dispenses. 

The wands and manifolds range between 4 and 48 tubes and are designed for microplates with wells between 48 and 1536.  This assay evaluated the performance of four separate devices, the VP 180, VP 184 (96 well format), VP 189L, and VP 189M (384 well format); devices that represent the fundamental design of the wands and manifolds.  The experiment also provided the data necessary to determine if the design differences between wands and manifolds affect the dispense performance.

Dispensing was done in a non-contact fashion (above the wells) and generally resulted in coefficient of variations below 5% for all devices tested.  It was also observed that despite the differences between the wands and manifolds in design, there were no apparent differences in the quality of dispensing capabilities.  Only four devices out of many different possible wand and manifold products V&P Scientific carries were tested.  However, this assay provides a proof of principle for the design of the wands and manifolds, thus similarly designed products are expected to yield similar results. 

Introduction

V&P Scientific carries a variety of row/column wands and manifolds designed for aspirating from and dispensing into microplates.  Manifolds are designed to have the inlet/outlet port located at the top center of the barrel while wands are designed to have to have the port located at the end of the barrel.  This allows a portion of the wand to act as a handle for ease of handling the device.  Both styles, however, have anywhere between 4 and 48 tubes designed for microplates with wells between 48 and 1536. 

The wands and manifolds can be connected to a vacuum line for aspiration or used in conjunction with a liquid dispenser such as a syringe, bottle top dispenser, or directly connected to an elevated source.  The focus of this paper will be regarding the dispensing capabilities of the wand and manifold using four different instruments representative of the fundamental designs; the VP 180, VP 184, VP 189L, and VP 189M.  The VP 180 is an 8-channel manifold for 96 format microplates, the VP 189M is a 16-channel manifold for 384 format microplates, and the VP 184 and VP 189L are the wand versions respectively. 

Most applications involving these devices generally do not require exact precision, such as aspirating media from wells or washing microplates.  However, some applications require that the instrumentation used to be precise and have a low coefficient of variation.  The precision of the wands and manifolds was investigated as well as observing any potential variation between the wands and manifolds for liquid dispensing due to design differences.

Background

The method of testing the wands and manifolds are fluorescence-based readings, using fluorescein isothiocyanate (FITC).  FITC is a derivative of fluorescein with the molecule functionalized with isothiocyanate, thus making it reactive towards amine and sulfhydrl groups, and a great tool for labeling proteins.  FITC has an excitation and emission wavelengths of approximately 495 nm/521 nm, and a maximum energy output at pH. 8.0. Because FITC is pH sensitive, it is important to keep it in a buffer to avoid even small shifts in pH¹.

The excitation value output of FITC is directly proportional to the amount of FITC molecules, hence a linear relationship between the amount of FITC and the excitation signal.  Using this model, the experiment is designed such that the amount of FITC dispensed by the manifolds in each well is read, thus each well can be compared to one another.  The percent coefficient of variation (%CV) can be extrapolated from this data using the formula:

The percent coefficient of variation is a recognized value in the laboratory for determining precision of instruments, thus allowing for comparisons to other equipments with similar functionalities (i.e. comparing the manifolds to multi-channel pipettors commonly found in laboratories)². 

Materials and Methods

• 8-Channel Manifold (V&P Scientific VP 180)
• 8-Channel Wand (V&P Scientific VP 184)
• 16-Channel Manifold (VP Scientific VP 189M)
• 16-Channel Wand (V&P Scientific VP 189L)
• Bottle Top Dispenser (V&P Scientific VP 195D-1)
• 96-well Polystyrene Black Assay Plate (Greiner Bio-One. 655076)
• 384-well Polystyrene Black Assay Plate (Greiner Bio-One 781076)
• Microplate Reader (Victor 3)
• Fluorescein 5-isothiocyanate, Isomer I (Sigma F7250)

·      Dimethyl Sulfoxide (Sigma D2650)

·      Tris-HCl, 1 M Stock Solution, pH 8.0 (Sigma T3038)

100 milligrams of FITC was dissolved in 4 mL of DMSO and left overnight to ensure complete dissolution and equilibration of the mixture.  Mix 80 uL of the FITC mixture with 2000 mL Tris-HCl, 0.1M, pH 8.0, buffer to get a final concentration of 0.001 mg/mL FITC solution.  This is the solution to be dispensed during the assay.

The final concentration to be used is a function of several factors, the most important to consider are the following; the readable linear range of the plate reader, the strength of the fluorescence, and the volume dispensed into the wells.  A standard curve was generated using a 12-channel pipette to determine the ideal concentrations and volume of FITC to be used, hence the concentrations used here reflects the equipment and reagents used here.

The wands and manifolds were set up according the manufacture’s technote.  The FITC solution was dispensed into 96-well microplates using the VP 180 and VP 184 at volumes of 125 uL and 250 uL per well.  Each volume was tested in three replicates (triplicate) resulting in a total of twelve individual replicates.  The VP 189M and VP 189L were tested at a volume of 62.5 uL per well.  Each device was tested in replicates of twelve resulting in a total of twenty-four individual replicates.  All dispenses were non-contact as the tubes neither touched the liquid nor wells. 

Results

Table 1 represents the data obtained using the VP 180 dispensing at 125 uL per well and 250 uL per well.  The CVs ranged between 3.19% and 9.75% for the 125 uL dispense, averaging 6.48%.  The CV for one of the replicates was particularly high due to one well (highlight) with significantly lower readings.  The average CV disregarding the highlighted cell would be 3.38%.  The CVs ranged between 2.35% and 2.82% for the 250 uL dispense, averaging 2.56%.

Table 2 represents the data obtained using the VP 184 dispensing at 125 uL and 250 uL per well.  The CVs ranged between 1.86% and 2.96% for the 125 uL dispense, averaging 2.46%.  The CVs ranged between 0.87% and 1.59% for the 250 uL dispense, averaging 1.26%.

Table 4 represents the data obtained using the VP 189M dispensing at 62.5 uL per well.  The CVs ranged from 2.52%% up to 4.58%%, averaging 3.70%.

Discussion

The VP 180 generally resulted in CVs between 2% and 3% with the exception of one replicate in Table 1 where the highlighted well significantly delivered lower volume than its peers (9.75% CV).  Delivery discrepancy can be due to mishandling of the manifold resulting in uneven dispensing (i.e. one side of the manifold is higher than the other) due to siphoning.  However, this particular well was most probably due to an air bubble caught in the tube dispensing into that well, as a siphoning effect would result in a high/low distribution of the data.  The fact that the under delivered well was in the middle of the manifold and that all other wells in that replicate represented average data, suggests that an independent air bubble was caught in the tube that dispensed into that well.

The VP 184 consistently resulted in CVs below 3% and as low as 0.87%.  Average overall CVs with a 125 uL/well dispense was 2.46% and 1.26% with a 250 uL dispense.

CVs for the VP 189L averaged approximately 4% consistently.  Again, one well here was significantly under delivered when compared to its peers as can be seen in the highlighted cell.  This was again probably due to an air bubble as observed in the VP 180.  The manifold version, the VP 189M also resulted consistently in CVs approximately 4%. 

Virtually no differences were observed regarding the CVs between the manifold and wand designs.  Both designs generally resulted in CVs below 4% using the 8-channel and below 5% using the 16 channel.  

Conclusion

V&P Scientific wands and manifolds can dispense into microplates while maintaining CVs below 5%.  Both wands and manifolds yielded near identical results, hence the design differences between the two do not affect the precision.  Only four devices out of many different possible wand and manifold products V&P Scientific carries were tested.  However, this assay provides a proof of principle for the design of the wands and manifolds, thus similarly designed products are expected to yield similar results.

References

1. W.C. Sun, K.R. Gee, D.H. Klaubert, R.P. Haugland.  (1997) Synthesis of Fluorinated Fluoresceins.  Journal of Organic Chemistry, 62, (19), pp. 6469-6475.
    

2. G.F. Reed, F. Lynn, and B.D. Meade.  (2002) Use of Coefficient of Variation in Assessing Variability of Quantitative Assays.  Clinical and Diagnostic Laboratory Immunology, Volume 9, No. 6, pp. 1235-1239.