LabTalk

figure 1

LabTalk is the built-in programming language that enables users to access all of Origin's features. Users have full control over graph and worksheet attributes, as well as the nonlinear curve fitter. In addition, manipulation of worksheet data is easy, and often requires only a few lines of code. Origin allows one to create worksheet and graph templates with graphical buttons (see figure 1). The commands needed to launch relevant scripts can be attached to these buttons (see sidebar) , creating a consistent analysis environment for all users.

All of the files used in this article are available for download in cyclase.zip. Included are the relevant template files (cyclase.otw, cyclase.otp ), the script file (cyclase.ogs ), and an Origin project file that can be used to initiate analysis (ACtemp.opj ). Installation of Origin 5.0 is required to use these scripts. These templates and scripts work with the demo version of Origin, available for download from Microcal Sofware.

The assay for adenylyl cyclase activity was conducted as described by Salomon [2]. Briefly, cells were labeled on 24-well tissue culture dishes with [³H]-adenine for 2 hours in the presence of 0.5 mM Ro 20-1724. Cells were washed, then exposed to a range of agonist concentrations in cell culture medium, then the reaction initiated with the addition of 10 µM forskolin. After stopping the reaction, each well was spiked with [¹⁴C]-cAMP. The same amount of [¹⁴C]-cAMP used for the spike was added to two empty liquid scintillation vials and used to calculate the efficiency of cAMP recovery. cAMP was isolated through successive columns of AG50W-X4 resin (Bio-Rad) and alumina (Sigma), and eluted with imidazole buffer directly into scintillation vials.

The counts per minute (CPM) for ³H and ¹⁴C channels were obtained from a liquid scintillation counter capable of dual-channel counting. Table 1 shows the raw data obtained in this experiment. The counts for the ¹⁴C channel (ch2) reflects the total [¹⁴C]-cAMP in the samples, and is used to calculate the % cAMP recovered using this column separation method.

Table 1: Raw CPM for [³H]-cAMP (Channel 1) and [¹⁴C]-cAMP (Channel 2). Duplicate data points were collected, and agonist concentration used for each of these points are presented in molar concentration

log M	3H (Ch1)	14C (Ch2)
-11	2573	1496
-11	2215	1312
-10	2559	1368
-10	2407	1380
-9	1282	775
-9	2085	1228
-8	1969	1344
-8	1685	1260
-7	1263	1492
-7	1368	1458
-6	1325	1356
-6	1206	1449
-5	1498	1094
-5	2023	1404
-4	2255	1223
-4	2381	1349
-3	2570	1389
-3	2058	1105

Manipulating Data

The raw CPM obtained for channels 1 and 2 are entered in the appropriate columns of the Origin worksheet. For this example, the simplest analysis is used; this ignores the subtraction of background counts for each channel, as well as the cross-channel counting that occurs. After entering the raw counts for each agonist concentration, one must simply click on the Run Analysis button to initiate the transformation.

The command attached to this button is run.section(cyclase,transform) ; this runs the script contained in the [transform] section of the cyclase.ogs file. This script is reproduced below:

First, the script prompts the user to enter the total [¹⁴C]-cAMP spike added to each well. In this experiment, the spike was measured as 2035 CPM. Next, the script iteratively calculates the efficiency of recovery for each data point (ch1), using the ratio of ¹⁴C counts (ch2) and the total spike (2035 CPM). Dividing each [³H]-cAMP data point by this efficiency results in a corrected CPM value that is entered into the Origin worksheet. An additional column is included to show that efficiency values range from 0.38 to 0.74 (See Table 2). Not shown in the above script are the commands to select and automatically plot the corrected data.

figure 2

figure 3

figure 4

figure 5

Results

Figure 2 shows the graph created after clicking the Run Analysis button on the worksheet template. The graph template includes another button that is linked to the nonlinear curve fitter. Clicking on this button selects the correct fitting function and opens the NonLinear Curve Fitting dialog box. This button runs the script in the [fitAC] section of the cyclase.ogs file.

The Biphasic logistic equation, (equation 1) found in the Pharmacology category of the nonlinear curve fitter, is used to fit this data.

The fitted parameters are Amin (minimum value), Amax1 and Amax2 (initial and final maximum values), x0_1 (log IC50), x0_2 (log EC50), and h1 and h2 (Hill coefficients for IC50 and EC50 respectively). Both h1 and h2 were fixed at 1 for this analysis.

Figure 3 shows the curve fitter with initial parameter estimates, and the curve that results from these values. Clicking on the "10 Iter." button starts the curve fitting process, which iteratively changes the parameters to result in a minimized chi square. An improved fit is illustrated in Figure 4.

Figure 5 shows fitted curves for raw CPM acquired from the liquid scintillation counter (red circles) as well as CPM corrected for recovery efficiency (black circles). The final parameter estimates obtained from this analysis is shown in Table 3. These results clearly show that raw CPM results in parameters with larger standard errors, and which differs from fitted parameters after correcting for column efficiency.

Table 3: Fitted parameter values for raw and corrected CPM

Parameter	Raw CPM	Corrected CPM
Min	1403 ± 145	1482 ± 121
Max1	2495 ± 192	3563 ± 56
Max2	2390 ± 197	3803 ± 74
log IC50 (M)	-9.20 ± 0.43	-7.74 ± 0.11
log EC50 (M)	-4.74 ± 0.49	-5.17 ± 0.10

EC50 and % inhibition for raw CPM and corrected CPM are not substantially different. Maximum % Inhibition (calculated using the relationship in equation 2) is 44% and 58% for raw and corrected CPM, respectively, and EC50 (10^logEC50) only differs by 3-fold. On the other hand, calculated IC50s (10^logIC50) are 0.6 and 18.2 nM for raw and corrected CPM respectively, representing a 30-fold difference.

Figure 5 includes one raw CPM data point marked in blue. This particular point results from low recovery efficiency at a value of 0.38 (see bold data in Tables I and II), and is very close to the duplicate data point once corrected.

Although the calculations presented herein are rather simple, one can easily increase the complexity of the analysis to correct for background counts and cross-channel counting. A few additional vials must be counted to collect this data, and one must write about 10 lines of LabTalk code to incorporate these corrections. Please contact the author (bulseco@os.com) for more details on these additional corrections.

Summary

The acquisition of high-quality data is difficult when one uses tedious, labor-intensive methods. Column purification is rarely 100% efficient, but analysis methods can be implemented that correct for recovery and dramatically improve data. In addition, under certain situations these corrections are essential to obtain the correct fitted values. This article demonstrates how Microcal Origin can facilitate data manipulation, plotting, and subsequent nonlinear curve fitting. Although many of these calculations can be accomplished in other spreadsheet programs, Origin enables one to create customized data analysis environments, making it easy for all users to analyze an experiment with a few mouse clicks.

System Requirements

Origin 5.0 from Microcal Software is available for Windows 95 and Windows NT 4.0 or later. A minimum of 8 Mb of RAM is required (16 Mb recommended), a PC with a 486/DX or higher processor, and a minimum of 8 Mb of available hard drive space.

Purchase Information

Origin 5.0 for Windows 95 and Windows NT can be purchased from Microcal Software for $595. Microcal Software can be reached by mail at One Roundhouse Plaza, Northampton, MA 01060, by phone at (800) 969-7720 or (413) 586-2013, by fax at (413) 585-0126, or by email. See Microcal Software for more information on Origin 5.0, or to download the demo version. If you already own Origin 5.0, be sure to download the Origin 5.0 Patch 2.

Dylan Bulseco is Research Associate at the Worcester Foundation for Biomedical Research and contributing editor of the HMS Beagle Software department.