ECIS Data Analysis for Stimulation of Human Pulmonary Microvascular Endothelial Cells (HPMECs) with Human Serum

Determine Difference in Area-under-the-Curve (Normalized Resistance at 4000 Hz)

To be calculated after recording full ECIS curve after stimulation with human serum from liver transplant (LT) patients.

Goal: To calculate the difference in AUC between the normalized ECIS curves generated by ( a ) Liver transplant serum and ( b ) Healthy pooled serum.

Using ECIS software, normalize the resistance. This is done by dividing the resistance at each time point by the baseline resistance (the plateau resistance reached just prior to stimulation with serum). This will set the baseline resistance = 1.

Export "Graph Data" and import into GraphPad prism.

Name data table " raw normalized_mm-dd-yy_cell line "

Duplicate X-Y tables with data from 0-6 hrs --> name duplicate " _adjustedt_0-3h "

Align time = 0

• Inspect normalized curves for a given experiment (day and cell line)
• Determine time point ( t _stim) where curves begin to diverge from baseline resistance = 1
• The time point prior to that time ( t _stim-1) will be set = 0 hrs
• Calculate time shift in Excel using the follwing formula: t _{adjusted =} t _{original -} t _stim-1* Use t _adjustedfor determination of AUC 0-3 hrs.

Delete data points with t _adjusted > 3 hrs

Analyze --> XY analyses --> Area under curve

1. Baseline Y = 1
2. Minimum Peak Height = ignore peaks < 5%
3. Minimum Peak Width = leave blank
4. Peak direction = "Also consider 'peaks' that go below the baseline."

From AUC Results --> Copy " Net Area " and " Std Error " to Excel.

Note: It is not necessary to copy/paste transpose each result individually. Copy the " Net Area " rows from the whole table and then use " =transpose() " function in Excel.

Calculate ΔAUCΔAUC = AUC LT serum - AUC Healthy pooled serum Notes: * Steps 2 and 2.1 can be performed in a single Excel sheet after running the "transpose" function.

• Both AUC and ΔAUC can be tabulated in a single Prism worksheet (Grouped)
• Store results in a new " Grouped " table in Prism (Mean ± SEM) named "AUC" with column labels: AUC, 0-3h ΔAUC LT - Healthy , 0-3h

Standard error calculation for ΔAUC

Std Error_ΔAUC ΔAUC = Where:

dx = Std Error_{AUC, LT serum}

dy = Std Error_{AUC, Healthy pooled serum}

Standard errors are obtained from the " Net Area " section of the Results tab in Prism after an AUC analysis.

Plot average curves with error bars from 0-6 h:

• Healthy serum with positive controls {LPS, TNF, IL-1b} and negative control {Media} - one for each day of experiment.
• Healthy serum with Liver serum {S1, S2, S3, S4, S5} - one for each patient .
• Generally there are 3 replicates, occasionally 2.

Upload mean ΔAUC_{0 to 3 h} in REDCap for each patient, timepoint, and cell line.

Note: Standard errors are available in Prism (too cumbersome for REDCap, and generally quite small).

REDCap variables for Cell Line #1 are named:

auc_ln1_s1

auc_ln1_s2

...

auc_ln1_s5

... similarly for Cell Line #2:

auc_ln2_s1

auc_ln2_s2

...

auc_ln2_s5

S1 vs. Healthy Serum - Q: Is liver serum at the start of surgery different from Healthy controls?

Perform one sample t-test (or one sample Wilcoxon signed rank test, if the normality assumption is not met) for each ΔAUC_{0 to 3 h} time point vs. 0.0 (the value for Healthy Pooled Serum).

S1 vs. S3 - Q: What is the effect of liver reperfusion?

Perform paired t-test (or Wilcoxon signed rank test, if the normality assumption is not met) for ΔAUC_{0 to 3 h}at the S1 vs. S3 time points, which were measured for each subject.

Perform additional pairwise comparisons for time points with missing data, as Supplementary Analysis:

Note: Control Column should be S1, Start of Surgery for all

1. S1 vs. S2 - Q: What is the effect of surgery (dissection phase) alone with no liver reperfusion?
2. S1 vs. {S4 or S5} - Q: If there is an effect of liver reperfusion, does it get worse with more time from reperfusion?

• S1 = Start of Surgery
• S2 = End of Dissection (before liver removal)
• S3 = 15 min after Portal Vein reperfusion
• S4 = 60 min after Portal Vein reperfusion
• S5 = 120 min after Portal Vein reperfusion (or End of Surgery, whichever comes first)

Cell line #* Passage #

• Old plate (yes/no)
• Seeding density (cells/well) - should be 40,000
• Plateau capacitance at 64,000 Hz, average and st. dev - prior to stimulation
• Plateau resistance at 4,000 Hz, average and st. dev - prior to stimulation
• Time of stimulation - elapsed from seeding ECIS plate