Overview
This page summarizes the key findings from our experimental analysis of IL-10 stability and outlines how these results guide our next steps. Our goal was to understand how environmental factors as well as engineered point mutations influence IL-10’s ability to remain stable and maintain its functional concentration.
Through statistical comparisons of experimental conditions, we identified significant differences that revealed where IL-10 is most prone to dissociation and how specific mutations can improve or reduce its stability. These insights not only validated our experimental approach but also highlighted areas where further testing and refinement are needed.
By consolidating these findings here, we aim to provide a clear picture of what we have learned so far and how it informs our strategy moving forward. The following sections present the major conclusions drawn from our data and propose realistic future directions to enhance reproducibility, optimize protein engineering, and strengthen the potential therapeutic application of IL-10.
Raw Data
Figure 1. Our raw data with labels pasted onto a google sheet & color coded for clarity. Two trials are run (in triplicates) for this ELISA plate, with the standard curve on the leftmost column and blanks in all right columns.
Two hours within adding the Stop Solution, we obtained the values listed above through reading the ELISA plate at an absorbance setting of 450nm. We color-coded each of the samples, distinguishing them by: 1) temperatures of 37C or 39C; 2) IL10 variants of original, mutation 1, and mutation 2; 3) solvents of lysate or supernatant; 4) concentration readings done in two different trial rounds. The leftmost column displays our standard curve, with the concentrations obtained through a serial dilution.
Data Analysis
First, we converted the absorbance values into concentration values to get more precise and universally comparable results. To do this, we used the AAT Bioquest Four Parameter Logistic (4PL) Curve Calculator to generate the standard curve. Optimally, the background absorbance may be subtracted from all data points, but the absorbance in all blank wells was 0 so no data was processed first.
In order to plot the standard curve, we put in the data points of the absorbance (or often referred to as optical density), and set up the calculator options to carry out background correction in order to subtract the values by the smallest response (0.042) since the value in the well with no Human IL-10 should’ve had 0 absorbance. We also normalized the data, which divided all values by the largest, in order to keep the values of absorbance between 0 and 1.
Figure 2. The standards used as values to calculate our curve are input into the software. The settings & parameters are listed above which include display of error bars, background correction, and normalization of values.
Figure 3. Our curve obtained through the standard absorbance data. Regression results are also displayed to show how to calculate additional absorbance values into concentrations.
We then used this equation to calculate the concentration for each well. However, as can be seen in the equation, the y-intercept was 0.0878, and any absorbance value lower than that would not be able to give a concentration. For those wells (highlighted in purple), we set the concentration as 0.
Figure 4. Restructured table of absorbance values. The values in purple could not yield a concentration as they were below the y-intercept value of 0.0878.
Figure 5. Table of calculated concentrations from the 4PL curve.Notice that the wells originally highlighted in purple were set to 0.
Data Analysis
We split the IL-10 variants into the two different trial rounds and a sum of both trial rounds. Using programming software R, we first graphed the concentration of the IL-10 variants into boxplots to visualize the data distribution and compare them. We also graphed a bar graph with error bars of 95% confidence intervals.
Figure 6. Boxplot of concentrations of IL-10 variants. On the x-axis are the types of IL-10 (temperature, lysate or supernatant, mutation number, and trial V or L). The trials with no labels of V or L are the average of the two. The y-axis marks the concentration.
Figure 7. Bar graph of mean concentrations of IL-10 variants with error bars of 95% confidence interval.
Statistical Tests
With R, we conducted a one-way ANOVA to compare the mean concentrations of all of the IL-10 variants, and the p-value of 1.69E-5. This is a very small p-value, signifying that we can reject the null.
Figure 8. One way ANOVA outputs. We can reject the null with a given p-value of 1.69E-5.
Since this only showed whether there is at least one pair that has a significant difference in its mean concentration, we conducted a Tukey’s Honestly Significant Difference (TukeyHSD) post-hoc test to see which pairs had the significant difference. However, none of the adjusted p values were <0.05, which we defined as statistically significant. We theorize that it likely has to do with our small sample size.
Thus, we conducted pairwise t-tests with an adjustment in p with the Benjamini-Hochberg (BH) procedure for all of the IL-10 variants. p<0.05 for t-tests between:
- 37L0 and 37S0
- 37L0 and 39S2
- 39L2 and 37S0
- 39L2 and 39S2
- 37S0 and 37S1
- 37S0 and 39S0
- 37S1 and 39S2
- 39S0 and 39S2
and more between the overall vs different trial rounds. For example, 37L0 and 37L2_L or 37L0 and 37S0_L. We did not deem these to have significant meaning so we did not count all of them. However, we have provided a comprehensive test between all different groups below as listed:
Conclusion
From the eight pairwise comparisons that produced statistically significant differences (p < 0.05), three central conclusions emerge regarding the behavior of IL-10 under varying conditions. These findings highlight how both environmental factors and engineered mutations influence the protein’s stability and binding efficiency.
First, our data demonstrate that temperature plays a measurable role in IL-10 dissociation. When the original IL-10 obtained from the supernatant was incubated at either 37 °C or 39 °C, we observed a significant difference in the concentration of IL-10 bound to its antibody. This pattern suggests that even a modest increase in temperature can disrupt IL-10’s conformational stability, reducing its capacity to maintain interactions with the antibody. Given that IL-10’s therapeutic effectiveness depends on its stability and bioavailability, these results underscore the importance of temperature control during both experimental assays and potential clinical applications.
Second, the results reveal that the source of the IL-10—whether collected from the lysate or the supernatant—significantly influences its concentration. For the original IL-10 incubated at 37 °C, samples derived from lysate exhibited markedly different concentrations than those obtained from the supernatant. A similar pattern was observed for the IL-10 variant carrying mutation 2 when incubated at 39 °C. This difference likely reflects how the cellular localization of the protein—either retained within cells or secreted into the surrounding medium—affects its solubility, folding state, or exposure to degradation. These findings highlight that not only the genetic sequence but also the source of the protein must be considered when assessing IL-10 stability.
Finally, our analysis indicates that engineered mutations can substantially alter IL-10’s concentration compared to the original protein. In particular, IL-10 with mutation 1 displayed a significantly different concentration from the original protein when collected from the supernatant and incubated at 37 °C. Likewise, IL-10 with mutation 2 differed significantly from the original when collected from the supernatant and incubated at 39 °C. These observations suggest that structural modifications introduced by these mutations may enhance or diminish IL-10’s stability under specific thermal conditions, thereby impacting its recovery and functional activity.
Taken together, these findings highlight the complex interplay between environmental conditions, protein source, and engineered mutations in determining IL-10’s stability and binding properties. Understanding these relationships is crucial for optimizing experimental protocols and guiding future therapeutic development of IL-10 variants with improved resilience to physiological stressors.
Future Directions
Building on what we learned this season, there are several realistic steps we could take to further improve the stability, reproducibility, and delivery of IL-10. These directions would allow us to strengthen both the scientific foundation of our work and its potential therapeutic application.
We could explore additional point mutations guided by computational tools such as Rosetta or ThermoMPNN to identify substitutions that improve IL-10’s thermal stability and resistance to degradation. We could also focus on the dimer interface—where IL-10 is most prone to dissociation—by introducing targeted mutations or engineered disulfide bridges. These incremental modifications could lead to more robust variants that are better suited to real-world conditions.
To better understand why certain variants behave differently, we could perform additional structural and biophysical analyses. Techniques such as circular dichroism, differential scanning calorimetry, and size-exclusion chromatography could help us monitor folding and aggregation. We could also use surface plasmon resonance (SPR) to measure how mutations affect IL-10’s binding to antibodies or receptors at different temperatures.
We could conduct future tests with more biological and technical replicates to reduce variability and strengthen statistical significance. Standardizing assay conditions—such as incubation times, sample handling, and ELISA workflows—would help reduce experimental noise. Regular calibration of instruments and the use of consistent internal controls could further improve reproducibility and confidence in the data.
Since we ultimately envision probiotic delivery, we could explore expressing and secreting IL-10 variants in Lactobacillus plantarum—our intended chassis—instead of focusing only on E. coli. Testing different signal peptides and secretion systems could also help us improve protein yield and stability in the supernatant, which was an important variable in our current experiments.
To move closer to demonstrating therapeutic potential, we could evaluate the IL-10 variants in cell-based assays that measure suppression of pro-inflammatory cytokines, such as TNF-α and IL-6. As a longer-term goal, we could explore testing these variants in gut organoid models to better understand their performance in a more physiologically relevant environment.
We could also integrate our protein engineering with more sophisticated delivery approaches. For example, coupling IL-10 variants with inflammation-responsive promoters in probiotics could ensure that the protein is secreted only during flare-ups. We could further examine protective encapsulation methods—such as alginate or other biocompatible polymers—to maintain protein stability during storage and transport.
In summary, these future directions focus on building a stronger foundation for IL-10 engineering while making our experiments more reproducible and clinically relevant. By combining computationally guided improvements, standardized testing, optimized production in probiotics, and innovative delivery strategies, we could make meaningful progress toward a practical, stable IL-10–based therapeutic for inflammatory bowel disease.