Update:
Objective 1: Evaluate improvement of sensitivity with TaqMan probes and finalize technology platform
In the previous year, we set out to explore dPCR or qPCR as technology platforms for an assay to enumerate soybean rhizobia (symbiotic Bradyrhizobium japonicum) in farmer’s soil. Work from FY22 culminated in a successful qPCR assay with a sensitivity limit of ~1000 rhizobia per gram (See FY22 report, Figure 1) (Assay Version 1.0). Development of an assay with a new technology platform, digital PCR (dPCR) which was advertised to have enhances specificity and sensitivity to qPCR, was also attempted. However, dPCR assays were unsuccessful due to high amounts of nonspecific signal in negative controls. Conversations with the manufacturer (QIAgen) indicated that incorporating TaqMan probes into the assay would overcome this issue in dPCR. TaqMan probes are a modification to amplicon-based molecular detection methods (qPCR/dPCR) that provide an added layer of specificity by binding to the amplicons and creating a detectable fluorescent signal when bound. Since TaqMan probes can also be utilized in qPCR and have the potential to increase sensitivity of our assay, we sought to evaluate incorporation of TaqMan probe technology into both qPCR and dPCR assays (Assay Version 2.0). Therefore, our first Objective of FY23 was to investigate the incorporation of TaqMan probes into qPCR and dPCR assays to both 1) finalize the selection of technology platform, and 2) evaluate improved sensitivity with their use.
TaqMan probe evaluation
Two approaches to TaqMan probe design were explored.:
1) A custom TaqMan probe assay was designed by ThermoFisher Scientific and tested. This assay however was ineffective and showed significant amplification in non-rhizobia control samples indicating a lack of specificity (data not shown).
2) We manually designed a TaqMan probe to incorporate with our previously successful primer sets utilized in the Version 1.0 qPCR assay (Figure 1).
To evaluate the accuracy of the new 2.0 TaqMan probe assay, we tested it using samples that were previously used to evaluate and calibrate the 1.0 qPCR assay. Overall, the 2.0 assay proved reliable and showed highly similar results to the qPCR assay with the same samples (Figure 2).
Next, we evaluated the new 2.0 TaqMan probe assay with dPCR using the same samples. While the assay was now successful using dPCR when the probes were incorporated, the sensitivity of the assay was vastly lower than qPCR (Figure 3). These data, combined with the higher reagent cost of dPCR lead us to finalize qPCR as the best platform for our assay.
Finally, we explored the capacity for increased sensitivity in qPCR using the new 2.0 TaqMan probe assay by using undiluted DNA samples from soil extractions. Previously, undiluted samples had too much background noise using the 1.0 assay and as a result samples needed to be diluted from ~100 ng/uL from the soil extraction to 10 ng/uL before the qPCR assay. Using the 2.0 TaqMan probe assay, undiluted samples were successfully detected without background noise and led to an enhanced sensitivity of the assay from ~1000 rhizobia per gram to ~100 rhizobia per gram.
Objective 2: Establish reliability using different soil types and sampling procedures, and optimize as necessary
We had two goals with evaluating the reliability of the NDSoy2.0 assay:1) evaluating the reliability of the assay with different soil types from the state, and 2) establishing the required sampling procedures to ensure the reliability of the assay.
Assay reliability across soil-types
To assess the reliability of the assay using different soils, we utilized previously assayed rhizobium-free soil from Central Grasslands REC (Central ND) and new soil from Williston REC (Western ND) that was expected to have limited rhizobium populations based on agronomic history in a rhizobium spiked soil experiment. Rhizobia were spiked into each of the soils in a 10-fold dilution series, DNA was extracted from each of the soils, and the soils were assayed with the NDSoy2.0 assay to estimate the quantity of rhizobia in each soil. The results from the experiment are shown in Figure 5A and 5B. While the central grasslands showed the expected linear quantification across the dilution, the soils from Williston deviated from a linear increase in quantification (Ct) as rhizobia concentration increased. This resulted in a significant underestimation of the rhizobia in Williston soil samples relative to Central Grasslands at high levels of rhizobia, but a similar quantification when low levels of rhizobia were present). Combined, the non-linear amplification across the assay as well as the reduced detection of rhizobia in Williston soil indicate that the starting soil may be an important factor to consider for bias when performing the assay, and for maximum accuracy across soiltyps. Sandy soils such as those found in Williston are recognized as creating DNA extraction challenges. Future efforts should be made to normalize the quantification to the starting soiltype either through optimizing the DNA extraction procedure or introducing an approach for normalization between soiltypes.
Robustness of Assay to Sampling Approaches
An assay to quantify rhizobia from farmers fields would be more useful if it were robust to sample handling similar to those that are used for soil chemical analysis. These differ substantially from sampling procedures routinely used for molecular analysis, in that they are often non-temperature controlled, may be shipped over several weeks and are dried prior to shipping. Whereas, routine protocol for molecular DNA analysis involves shipping on ice over 24 hours and immediately freezing before DNA extraction is performed. To evaluate the robustness of the assay to less rigorous sampling approaches that may be used for soil chemical analysis, we compared three field soils from previous soybean fields from Hettinger, Williston and Carrington RECs. The soils were shipped on ice from the RECs to NDSU following sampling, and either immediately frozen, or dried at room temperature for two weeks prior to extracting DNA and comparing rhizobia quantification. Overall, we did not observe significant differences between the two sample processing procedures, indicating that the assay is robust to different approaches to sample handling prior for soil chemical analysis, including drying the soil and leaving it at room temperature for several weeks.
Objective 3: Test finalized assay using farmer’s field soil, with a focus on inoculant survival in acidic soils from Western ND.
As a final objective of the FY23 grant, we aimed to continue testing of Farmer’s field soils to gain a more robust understanding of rhizobia populations across North Dakota. FY22 data indicated significantly lower populations of rhizobia in field soils in Western ND. We considered this may lead to requiring different inoculant recommendations than those established primarily based on studied done in Eastern ND. Therefore, further data collection from fields across the state was carried out to further investigate this phenomenon. The data showed a variety of rhizobium population levels that varied from not detectable to very high populations (Figure 7). There was little difference between Western and Eastern ND when 2022 and 2023 data was taken together and rhizobium population level was compared to years since the previous soybean crop (Figure 8A). Overall, the data indicate congruence with current recommendations that rhizobium populations remain high in fields until ~the fifth year since soybean planting (4 years since last soybean crop at time of measurement) (Figure 8B) and our data support continuing the recommendation to inoculate on the fifth year since previous soybean crop. Our assay suggested tentative inoculate recommendations primarily for fields either without a history of soybean planting, or over 5 years since the previous soybean crop. However, we did identify one field in Eastern and one field in Western North Dokata with non-detectable levels of rhizobia despite only 2 years since the previous soybean crop (** in Figure 7, Figure 8B). This indicates that our tool to quantify rhizobium populations could be useful as an insurance case for scenarios when the typically expected dynamics of rhizobium populations don’t hold up for a given field, risking poor nodulation. Before implementation of the tool, data should be gathered for fields with high and low populations of rhizobia to assess the impact of inoculation on nodulation and yield. By connecting inoculant field trials with the results from the NDSoy2.0 assay in FY24 we hope to near implementation of the assay as a valuable agronomic tool to monitor rhizobium populations in the soil, and as a tool for farmers to assess the need to inoculate in a coming year.
View uploaded report 
View uploaded report 2
Unnecessary inoculation wastes farmers’ money and cuts into their bottom line. However choosing not to inoculate carries significant risk; if optimal nodulation does not occur, soybean crops may not get enough nitrogen resulting in yield losses. Unfortunately, no practical approach currently exists for farmers to predict the requirement for inoculation prior to planting. This work continues efforts to develop a cheap and reliable tool to track rhizobia in farmer’s field soils, informing inoculant decisions for farmers either indirectly through accelerating agronomy or directly through use a service. In research this funding year, we showed we could improve the sensitivity of the assay we previously developed by incorporating fluorescent probe technology. Next, we investigated biases in the assay when utilizing vastly different soils from Central Grasslands REC and Williston REC. We found that sandier soils from Williston did result in an underestimation of the number of rhizobia present, and identified improving the robustness of the assay to different soiltypes as an area for improvement in future years. Excitingly, we did find the assay proved robust to diverse sampling procedures. While careful preservation at cool temperatures and freezing of samples is typically thought of as required for molecular analysis, we found our quantification was nearly identical when soil samples were dried and stored at room temperature for weeks; conditions that might be expected for samples sent by farmers for soil chemical analysis. These data indicate the viability of this tool as a possible “add-on” to soil chemical analysis for farmers as they test their soils in the Fall or Spring before planting. Finally, by analyzing 23 different field samples and combining the data with a similar number of field samples from 2022, we began to get a picture of how rhizobial populations shift in North Dakota farm fields. Although reduced populations were expected in Western North Dakota based on 2022 data, with the expanded dataset from 2023, less differences were observed. Across the state, populations generally remained high up to 3 years after soybean planting and declined after that which is roughly in line with state recommendations to inoculate every 5 years. Some fields as little as two years post planting however were found to have undetectable levels of rhizobia. This indicates the potential value of the tool as a failsafe to indicate if farmers should inoculate sooner should their fields not follow typical expected patterns of rhizobia/nodulation over time.