Koch's Postulates on Hop Latent Viroid

Part 2

Figure 1. The Medicinal Genomics Lariat assay can discern latent HpLVd from actively replicating HpLVd.

This is a running Lab Blog on a series of experiments to assess if HpLVd infection of Cannabis can satisfy Koch’s Postulates

To recap-

Collin Palmer’s UK Cheese plant tested positive in the leaves for HpLVd during vegetative growth back in September, 2023. After flowering, this plant turned deep purple (anthocyanin production) and then proceeded to test negative in the canopy ~4 weeks later on purple leaf tissue.

Why do we care about this anthocyanin production? Xu et al performed transcriptome analysis with Hop Stunt Viroid infection in sweet cherry and found genes involved in anthocyanin production up-regulated. This is not the same viroid as HpLVd but it may be predictive of plant responses with other related viroids.

UK Cheese being HpLVd positive in the leaves during vegetative growth and then turning HpLVd negative in the anthocyanin rich leaves during flowering is a curious finding. This led us to get access to the roots to see if the plant had cleared the viroid entirely or merely constrained HpLVd to the roots as we have seen in the past with Jamaican Lion cultivars?

The Roots were in fact HpLVd positive and the green tissue in the canopy is also positive but at higher HpLVd CT than the roots (higher CT = less HpLVd). Purple tissue continues to test HpLVd negative (>CT35).

Figure 2. qPCR curves of HpLVd and Internal Cannabis Controls

This was first observed with Jamaican Lion (JL) last year. I presented at CannMed on a large multi-tissue RNA-Seq study we performed on Jamaican Lion through the course of an HpLVd infection. This cultivar was never positive in the leaves. After injection of a ground up leaf slurry from an infected vegetative plant, JL only became HpLVd positive in the roots. This mode of infection isn’t a purified viroid and any symptom transmission could not be solely attributed to HpLVd transmission.

To complement the UK Cheese data, we infected Jamaican Lion clones with purified HpLVd circRNA (synthesized). It is important to emphasize that the molecules were not ‘isolated’ from a mixture of other biologicals but were chemically synthesized ground up from scratch with no other biological contaminants. There is no “it has never been isolated” canard with synthetic genomes.

These Jamaican Lion clones became positive in the roots 5 days post infection (dpi) and were negative in the canopy at 5dpi.

We injected 723 billion to 7.23 trillion circRNAs into the stems of 2 JL clones. This large quantity of injection could simply manifest as a dilution effect into the plant given the CT scores in the roots are lower than what was injected. The canopy is HpLVd negative 5dpi suggesting an uneven dilution effect of the injected viroid.

At 5dpi the Jamaican Lion clones were switched to 12 hour lighting (flowering).

Figure 3 - Jamaican Lion 5 days after injection is positive in the Roots but negative in the leaves.

After 5 days of flowering (10dpi), HpLVd can now be detected in the leaves, petioles and meristem with RT-qPCR.

Table 1. HpLVd tropism in cannabis plants

So how do we discern simple dilutions of the circRNA from replicative RNA?

First you need to know a little bit about how these viroids replicate in the host. They are too small (256bp) to code for any polymerases so they must rely on host replicative machinery.

HpLVd uses an Asymmetric rolling circle amplification, where the circular RNA template is replicated by a host polymerase into linear concatemers depicted in blue. These are used as templates for RNA polymerases, RNases and Ligases to make further circles which repeat the process. The exact enzymes that perform this replication in Cannabis have not been identified but the genome sequencing of Jamaican Lion has created a list of candidate orthologs to plants where these processes are better understood.

Figure 4- From https://encyclopedia.pub/entry/2834

In order to understand the replicative potential of injected synthesized circRNA, monitoring the linear to circular RNA ratio becomes important. The input synthetic circRNA will have a constant circular to linear ratio. Any circRNA that is actively replicating will have a skewed circular to linear ratio.

To address this, we developed a Lariat assay to ascertain the replicative stage of the viroid. This leverages RNaseR which only degrades linear RNA and cannot hydrolyze circular RNA.

Figure 5. The Medicinal Genomics Lariat Assay can discern Active HpLVd from Latent HpLVd.

By running this assay on the injected synthesized circRNA and comparing it to the RNA found in the plant we can ascertain if the injected viroid is replication competent. There should be linear RNA forming after infection that is detectable with RT-qPCR.

First we survey this assay on the biologically infected UK Cheese to understand the linear to circular variation compared to the synthesized control.

Figure 6 UK Cheese roots after harvesting. These roots were very dry and the plant was many weeks past flowering with necrotic tissue beginning to emerge.

Figure 7. No RNaseR Cq is subtracted from RNaseR. 2 to the power of this delta CT is plotted above to estimate the ratio of Linear to circRNA ( 2^(Delta CT))

The control sample at the bottom of Figure 7 is pure circular RNA. About 5% of this is believed to be linear according to the GeneScript QC. All the samples with larger Cq offsets than the circRNA control have more linear RNA than circular RNA demonstrating replicative HpLVd in the plant. This half CT difference on the circRNA control performed with and without RNaseR is likely within the error of qPCR and pipetting variances. ddqPCR is needed to better quantitate less than 1 CT differences.

UK Cheese is a naturally infected plant and serves as a biological baseline for what we should see in the injected plants if the viroid replicates.

As we turn to the infected JL plants that are 5 days into flowering and 10dpi we see much more linear RNA being synthesized by the plant. The linear to circular ratio is largest in the canopy.

Figure 8. Cq scores for Jamaican Lion clones infected with synthetic HpLVd circRNA. RNase R and zero RNaseR samples are plotted in Blue and Red respectively.

Figure 9. Delta Cq for treated and untreated samples show more linear circRNA in the canopy.

This is an interesting observation. At 5dpi the viroid was only detectable in the roots implying simple dilution through the plant had to concentrate the viroid in one tissue. After 5 days of flowering, we can see the viroid has moved to the canopy and has a much different linear to circular RNA ratio than the roots. The CT values in the canopy are also lower than in the roots implying more HpLVd genome is being generated. This negates an alternative hypothesis of plant RNase activity simply nicking the circular RNA as it moves throughout the plant and thus creating RNA that is RNaseR sensitive.

To assess the sensitivity of the assay, both Limit of Detection analysis (LODs) and Deep RNA sequencing was performed.

8 additional biologically infected plants (different cultivars) that tested positive with this RT-qPCR assay were RNA sequenced at Medicinal Genomics with Illumina. Over 100M reads were sequenced from each sample. The viroid was detected in 8/8 samples at less than 1 read per million. Note the viroid has no PolyA tail so its capture efficiency in this process is likely very low but the sequence information confirms CAN1 and CAN2 have been sequenced confirmed with this assay.

Figure 10. RNA-Seq of HpLVd infected plants. Over 100 million reads per sample generates ~100X coverage of the HpLVd genome.

Figure 11. IGV view of the HpLVd genome shows low variation in HpLVd sequences.

DNA positive controls were serially diluted to construct a DNA based LOD. This was performed on Version 2 build of the MGC assay that has been optimized for Root extraction.

Figure 12. 10ul Reactions were constructed in a MIC/MYRA automated qPCR station. Serial dilution of a positive control demonstrate 10X more sensitive LODs with the MGC version 2 assay in difficult to amplify root samples.

Hairpined CircRNA LODs are 2-3 CTs offset (later/less efficient) from their DNA LODs. Independent of this, the Version 2 assay is 10X more sensitive in Roots than our previous HpLVd assay. This is largely a result of the magnetic lysis step used to clean up dirty root samples.

Conclusions

Synthetically synthesized CircRNA can be injected into the stem of Cannabis plants at 723 Billion to 7.23 Trillion molecules and 10 days later demonstrate a linear to circular RNA difference in Cq scores with RT-qPCR and RNaseR treatment. Given these CT scores are lower (more RNA) than the original CT scores detected in the roots at 5dpi and the linear to circular RNA ratio is altered compared to the roots, this is highly suggestive of replication of the synthetic genome post infection. This partially supports Koch’s postulates as this synthetically synthesized and pure viroid is transmissible and replicative.

What has yet to be determined is if this transmissible viroid also confers disease in these cultivars as they have not matured to final flowering. The association with low yield has been made many times before but in studies that could not account for the vast array of other pathogens that may have co-associated with the transmission event. A very informative read on this topic recently published from Zamir Punja which demonstrates the pathogenicity of HpLVd appears to be cannabis cultivar dependent.

Final Sanger sequencing of the plant replicated viroid will be performed to verify the fidelity of the primers. Previous Sanger sequencing of viroids detected with these primers have provided 99% identical HpLVd sequences known as CAN1 and CAN2. These sequencing results are public on Viropedia.net.

The qPCR assay has been validated with Illumina RNA-Seq and can detect HpLVd viroids at less than 1 read per million of the mRNA in the sample (Billions of reads surveyed) or 5.5 copies per reaction.

The LODs for this assay has been calibrated on circular RNA templates and synthetic DNA templates. Since synthetic DNA templates are not hairpins and do not require a low temperature Reverse Transcriptase step, they provide an optimistic assessment of the true biological LOD. Synthetic DNA templates thus exaggerate your LOD 2-3CTs (10 fold), leading you to believe you have a much more sensitive assay than you really have once deployed on real circRNA templates.

For samples that are actively replicating and have more linear RNA than circRNA this distinction may be irrelevant as the linear RNA is not expected to be as difficult to denature and amplify as the circRNA. For latent samples where the RNA is mostly circular this difference may affect the assessment of the sensitivity of the test. As a result of the sequencing depth and synthetic RNA controls, we believe this is the highest sensitivity test on the market for HpLVd.

Understanding the viroids latency and replicative state may assist in breeding for cannabis plants more tolerant to HpLVd infection.

Some of the details on the RNaseR methods below-