Home | Sign Up For Our Newsletter

CASE STUDIES

Should I Submit Mouse Serum or Plasma to Rules-Based Medicine for Rodent Multi-Analyte Profiling?

Introduction

When discussing projects with our customers, they often ask, "Should we collect serum or plasma in our mouse study?" In this case study, we used data generated by our Rodent MAP antigen panel version 1.5 to compare the mean values for each of the 59 analytes measured. These data were generated with 3,502 mouse serum samples and 2,946 mouse plasma samples submitted to RBM for testing (386,880 data points). These data therefore represented many different strains of mice across many different experimental regimens. The purpose was simply to answer the question: "Is my analyte of interest better represented in serum or plasma?" This may serve as a guide for your own experimental design.

Methods

Data sets generated with mouse samples between June 2004 and June 2006 were utilized. The indicator analyte was fibrinogen as high fibrinogen levels are found in plasma samples and low fibrinogen is indicative of serum. Any data set with ambiguous fibrinogen levels or mixtures of serum and plasma were excluded in the data mining exercise. Only EDTA plasma sample data was used. Heparinized or citrated plasma samples were rarely found in these data sets and they were excluded. The mean value for each analyte with each sample type was computed and the mean compared with the Least Detectable Dose for that particular assay. The LDD was defined as three standard deviations above mean background. Those analytes whose mean value fell below the LDD were not included in the final analysis. Finally, for each analyte, a ratio of the serum mean value to the plasma mean value was shown in tabular form and also visualized using a bar graph.

Results

The tabulated results are shown in Table 1. The ratio of serum analyte value to plasma analyte value was listed in the last column on the right. Light green shading identified those analytes whose ratio of serum to plasma mean value was greater than 1.5. These analytes were significantly better represented in serum than plasma. Light blue shading highlighted those analytes whose ratio was less than 0.5. These analytes, Fibrinogen and von Willebrand’s Factor, as expected were better represented in plasma than in serum. Only those analytes whose mean value was above the assay LDD were noted. There were also some analytes with mean values below the LDD that also displayed greater representation in serum than in plasma. Due to the large number of samples used, those may be real differences between serum and plasma as well. Tables 2 And Table 3 list the two groups of analytes described.

Conclusions

When comparing the mean analyte values in mouse serum and plasma, it was apparent that there were many more analytes better represented in serum. A total of 20 analytes had a serum to plasma ratio greater than 1.5 where as only 2 analytes demonstrated a ratio less than 0.5. From these data, it appeared that serum was the best mouse sample when examining those 20 analytes. However, could it be that the clotting event itself caused much of the analyte elevation in serum vs. plasma? If there is interest in fibrinogen or von Willebrand’s factor, then plasma would clearly be a better sample. Because the sample number was so high in this study, we are quite confident in the outcome. Serum has always been the more common rodent sample type submitted to RBM and our customers tell us that they find it easier to collect serum consistently rather than plasma. As we have always stated, the most important factor in sample collection is that all samples within a study should be collected and the serum or plasma harvested and stored in a consistent manner. We present this data as a case study and you can decide which sample type meets your needs.

In subsequent newsletters we will address the rat serum vs. rat plasma data from a similar study as well as the rat vs. mouse reactivity with the Rodent MAP version 1.5 assays. In the fall of 2006, we added 9 new analytes in the conversion of the Rodent MAP to version 1.6. We will commence a similar study on those 9 analytes later this year when the sample number tested with the new version reaches a sufficient number.

We want to re-emphasize that this study included many different mouse strains, disease models, and experimental groups. Please do not assume that these data can be substituted as control data for a mouse experimental protocol. As we have stated previously, it is important in your study to submit gender and aged matched control samples from animals housed in the same facility as the experimental group. Only then can you glean the robust biomarker patterns that are commonly determined using the RBM MAP approach.

Please feel free to contact us with any questions or comments at info@rbmmaps.com.