Yeast and Mold Assay Method Validation

We have modeled our method validation for total yeast and mold after the AOAC SMPR for Aspergillus, AOAC Appendix J, and we are participating with the AOAC's Emergency Response Validation of TYM.

Updated October 2020

In 2019, the AOAC developed a Standard Method Performance Requirements (SMPR) for Aspergillus, E.coli, and Salmonella detection on cannabis. These are three separate SMPRs.

AOAC does not yet have an official SMPR for Toal Yeast and Mold (TYM), but we can take what we have learned from this process on Aspergillus and apply it to our TYM validation.

Our original validation was completed in 2015 when we launched the assay. The data below reflects current validation work performed in our lab in 2020 and we will continue to update this page with data.

As of October 2020, Medicinal Genomics will participate in the AOAC's Emergency Response Validation for the enumeration of viable total yeast and mold in dried cannabis flower. This project was initiated by several labs in Michigan and regulators there challenging the validity of Cq to CFU conversions. This new validation will be thorough and take months but will result in an AOAC certified performance tested method validation for TYM.

Total Yeast and Mold Assay Design

The PathoSEEK® Total Yeast and Mold Detection Assay is used to detect species classified as yeasts and molds. The results of a Total Yeast and Mold (TYM) test are an indication of yeast and mold contamination on a cannabis sample. The Assay targets these species using the FAM fluorophore. 

Regulations for total count microbial threshold tests currently use nomenclature from culture-based methods to set acceptable thresholds. It is necessary to convert the quantification cycle (Cq) value from the qPCR assay to colony forming units (CFU). In order to maintain the best possible concordance with plating, an imperfect method, the PathoSEEK® Total Yeast and Mold Assay does not include certain species that fail to culture. Powdery Mildew and Botrytis are two such species and are therefore omitted in the assay design. 

Peer-reviewed studies have shown that qPCR testing methods, such as the PathoSEEK Microbial Safety Testing Platform, are able to more accurately detect and quantify yeast and mold species present on a cannabis sample than culture-based methods. The PathoSEEK Total Yeast and Mold test looks for a specific DNA sequence that exists in all yeasts and molds. Because of this, the PathoSEEK Total Yeast and Mold test can detect all yeast and mold species, regardless of whether they grow in culture. Culture-based tests can only detect the species that grow in a given medium and timeframe. Conversely, qPCR can more effectively exclude off-target microbial species, such as bacteria, which have been shown to grow in TYM culture medium.

For more information about Cq/CFU conversions:

https://help.medicinalgenomics.com/what-are-cq/cfu-conversions

Inclusion and Exclusion Data

The first step of this process is qualitative inclusion and exclusion data for the primers. The below table contains our most current data, and may be more up to date than our published method validation document. Downloadable here: https://www.medicinalgenomics.com/validation-documents/

October 2020 Update: Once we are complete with our updated method validation work we intend to publish individual validation documents for each of our assays.

A minimum of 50 inclusion organisms and 30 exclusion organisms were selected according to the AOAC SMPR and FDAs general guidelines for the validation of qualitative detection methods for microbial analytes. We're still working through this list and have more to add, but our work to date is listed here. All ATCC#s listed are live samples unless they have a D in the name (it is a DNA sample).

Note: Data for the Total Aerobic Count assay is included in this table as well.

Table 1 - Inclusion and Exclusion testing of MGC TYM assay

Results

We detect 104/106 (98%) Yeast and molds with MGC TYM qPCR. Rhizopus stolonifer failed to detect. Rhizopus stolonifer also did not grow on 3M YM after 6 days of culture at Room Temperature (Figure 2). 2/3rd of the Aspergillus orzyae were detected. We are sequencing the 3rd A. oryzae to determine if this is a sample swap. Botrytis cinerea was not detected but as mentioned above this does not form colonies in 48 hours and will create discordance with CFU/g correlations. As a result, we have made an independent assay for both Botrytis cinerea and Golovinomyces chicoracerum (Powdery Mildew also doesn’t culture) so these microbes can be independently quantitated without complicating concordance with CFU counts. This represents over 98% detection of yeast and molds. There are 37 bacteria in the exclusion set and 37/37 (100%) bacteria were not called. Thus, we observe 98% inclusion and 100% exclusion.

For more information about these assays please visit our pathogen detection applications page: https://www.medicinalgenomics.com/pathogen-detection/

This is better than the published 3M data seen in Figure 1 (below).

Figure 1- Inclusion/Exclusion Table replicated from 3M YM product description. Note 5/19 fail to grow in 3 days and most cannabis labs are running 48 hour plates. Also note 4 Bacteria grew on the YM plates.

Quantitation Calibration

18 Yeast and Molds were selected based on their use in the AOAC Aspergillus SMPR or their published presence on Cannabis (Punja et al). Organisms were ordered from ATCC. 6 of these organisms have quantitative controls also sold from Microbiologics (https://www.microbiologics.com/) (See Figure 2 below)

Table 2- List of organisms to be used for correlation of CFU to Cq.

The organisms in Table 2 (above) are being grown on 3M RYM and PDA plates from Teknova without selection.

Figure 2. Microbiologics Aspergillus brasiliensis controls.

Aspergillus brasiliensis

540,000 CFU’s/ml were plated and run through the TYM qPCR assay

1:10K and 1:100K dilutions are countable on 3M RYM plates (Figure 4).

1:10K plates show 210,000 CFUs/ml on 3M RYM plates.

Enzymatic Lysis step for qPCR included.

Expect Ct in the 18-20 range.

Figure 3- Aspergillus brasiliensis plated on 3M RYM. 1 Disc was homogenized in 10mls and 1 ml plated with various 10 fold dilutions. 1:10K dilutions have 21 colonies and back calculate to 210,000 CFUs. This should generate Cq’s in the 18-20 range. 

Note: It is difficult to see the text in the above images. This just serves to demonstrate what our R&D group is actively pursuing and showcase some early results. The data entered in to our published method validation will consist of the proper number of replicates per AOAC guidance.

Figure 4- Aspergillus brasiliensis qPCR with MGC TYM assay. Various dilutions are displayed (Table 3 below). A new enzymatic lysis step (37C for 30minutes) is incorporated to manage beta-glucan and chitin rich cell walls found in C.albicans. Cq’s are within expected range (Cq=20.01) from Microbiologics standards.

Note: Once again this data is to show our work on further validation of the TYM assay. Future published method validation updates will be consistent with AOAC guidelines as per the Emergency Response Validation for TYM.

Table 3- Exploring Lysis Robustness. Various different Enzymatic lysis conditions were explored compared to a control methods (red). A 1.3 Cq gain is observed with Enzymatic lysis. The control method provides a Cq of 20 which is in line with expectations from Microbiologics control.

Aspergillus brasiliensis was repeated on PDA without selection.

Figure 5. Microbiologics Aspergillus brasiliensis plated on PDA without selection in duplicate across multiple 10 fold dilutions.

Figure 6. Quantitative qPCR results with A.brasiliensis demonstrates detection down to 56 CFU. Calculator may need slight adjustment once matrix is tested.

Table 4. Aspergillus brasiliensis plating results on PDA in duplicate compared to MGC.

Penicillium citrinum

Figure 8. Penicillium citrinum plated on 3M YM plates.

Table 6. Penicillium citrinum plated on 3M YM and compared to MGC TYM. qPCR is over counting compared to Aspergillus brasiliensis.

These results will be repeated on PDA without selection.

AOAC Emergency Response Validation and the path forward for Medicinal Genomics

We are delighted that the AOAC has stepped in to manage the validation of total yeast and mold testing for cannabis. We have great confidence in the AOAC and are members of the Cannabis Analytical Science Program (CASP) and will be contributing to the Emergency Response for TYM. Please see AOAC's page for more detail:

https://www.aoac.org/news/aoac-international-activates-accelerated-program-to-evaluate-test-kits-that-measure-yeast-and-mold-in-cannabis/

There has been a lot of misinformation and confusion in the field with regards to qPCR and enumeration for TYM on cannabis. There has been equally as much confusion and misinformation about what techniques are actually certified by any agency or governing body for cannabis testing. None!

Our next steps will be to validate our method according to their guidance developed in the emergency response by the AOAC. In parallel we will continue to build out our inclusion and exclusion criteria and compare results with plating to show correlation between the two.

We will be repeating the serial dilution of microbes on 5-day PDA plating without selection and for 15 more organisms. We expect this to be a fairly straightforward correlation to current CFU/Cq curves.

Part of AOAC Appendix J is asking for spiking these controls onto cannabis matrix with background microbiota. Cannabis with background microbiota will have bacteria that will grow on PDA without antibiotics and we expect this to complicate the CFU/g to Cq correlations as the bacteria will grow faster on the plates than the mold and CFU are very likely to include bacteria. PDA with Tartaric acid was recommended for this. This plan may change once the AOAC finalizes its design plan for the Emergency Response Validation.

We will continue to update our progress here. 

Thank you,

The Medicinal Genomics Team