Comment on “Evolutionary transitions between beneficial and phytopathogenic Rhodococcus challenge disease management”

We would like to address a number of concerns regarding this paper (Savory et al., 2017)

While the analysis by Savory et al. (2017) of Rhodococcus fascians isolates that are found to cause leafy gall disease on ornamental plants is noteworthy, the data presented do not unambiguously demonstrate plant growth promotion by PBTS Rhodococcus isolates. In contrast, the data presented in Savory et al. (2017) (Figure 7B and Figure 7-figure supplement 1) indicate that, upon inoculation, the Nicotiana benthamiana roots were stunted in comparison to the roots of control plants. PBTS Rhodococcus 2 caused significant root stunting at an inoculum level of only 2.5 Â 10 6 CFU ( Figure 7B), 10 3 fold lower than the standard dose used throughout the paper. In orchard systems, inadequate root development affects tree anchorage in soils and influences aboveground tree and canopy development. For these reasons, the data presented by Savory et al. do not support the assertion that PBTS Rhodococcus isolates are plant growth promoting.
Potential misdiagnosis of an epidemic Savory et al. claim to have 'uncovered a misdiagnosed epidemic' that has 'perpetuated the unnecessary removal of trees and has exacerbated economic losses in the pistachio industry'. The trees that exhibited Pistachio Bushy Top Syndrome (PBTS) were removed due to their irregular growth and development, and their high percentage of grafting failure (>70%; Stamler et al., 2015a). Those rootstocks that did successfully accept a scion bud developed bark cracking along the bud/graft union that weakened the tree. Pistachio ranch owners and managers of affected orchards concluded that tree removal was the only economically feasible option well before Stamler et al., 2015a was published. An ongoing study has demonstrated that PBTS trees entering maturity exhibit a 73% reduction in yield and lower nut quality in comparison to asymptomatic trees of the same age in the same orchard . Ranch managers opted for the more expensive hand-harvest, as opposed to mechanical harvest of symptomatic trees, due to the risk of breakage at the cracked graft union. Three of the authors on Stamler et al., 2015a are pistachio horticulturists with close connections to the pistachio industry. We are personal witnesses to the fact that the statements regarding tree removal in California, Arizona and New Mexico are inaccurate (Savory et al., 2017). The majority of PBTS symptomatic trees were removed from orchards a year before Stamler et al., 2015a was published (Northcutt, 2015;Fichtner et al., 2015). Hence, the claim that the trees were removed due to a misdiagnosis is erroneous, unjustly discredits the research team involved in the diagnosis, and insults the pistachio growers who decided to remove trees.

Presence of virulence genes
Savory et al. state that 'contrary to previous findings, there is no evidence to support the diagnosis that Rhodococcus without a virulence plasmid are responsible for an unusual growth problem that has plagued the pistachio industry.' It was never claimed that PBTS was caused by Rhodococcus isolates lacking virulence genes (Stamler et al., 2015a). On the contrary, the isolates utilized for establishing Koch's postulates were shown to have virulence genes (fasD and fasA) using PCR amplification and direct sequencing (Stamler et al., 2015a). Since previous work on R. fascians isolates from ornamental plants has shown that the presence of virulence genes is required for pathogenicity (Desomer et al., 1988;Crespi et al., 1992;Eason et al., 1995;Eason et al., 1996;Stange et al., 1996;Nikolaeva et al., 2009), we agree that PBTS isolates without these described virulence genes are likely non-pathogenic.

Subpopulations and loss of detectable virulence genes in PBTS Rhodococcus isolates when culturing on synthetic media
Savory et al. use a 'Genome Announcement' (Stamler et al., 2016) as evidence for the lack of virulence genes in PBTS1 and PBTS2, although the title of the cited paper clearly stated that the information provided concerned draft sequences of PBTS1 and PBTS2, thus recognizing that not all of the genetic information was obtained (i.e., fasD, fasA, and attA). Our unpublished work has shown that these specific Rhodococcus isolates obtained from UCB-1 pistachio trees and grown on synthetic media contain a subpopulation of bacterial cells with virulence genes (fasD and fasA). After subculturing the bacteria only twice on synthetic media, these virulence genes become undetectable by PCR. Thus, when sequencing these subcultured isolates using the PacBio platform, the virulence genes (originally detected using PCR amplification) were no longer detected (Stamler et al., 2016). Plasmid instability and loss of virulence are recognized in several bacterial species and were also described for leafy gall-inducing R. fascians isolates. Sabart et al., 1986 demonstrated that R. fascians (previously named Corynebacterium fascians) lost virulence within several generations due to growing conditions, and Nikolaeva et al., 2009 showed instability of virulence due to plasmid loss in certain R. fascians isolates.
High plasmid instability in R. fascians Savory et al. state that Nikolaeva et al., 2009 used the wrong molecular tools to evaluate pathogenicity, but these authors showed that for some isolates, individual fas positive colonies gave rise to fas positive and fas negative bacteria, very effectively ruling out the possibility that these findings were caused by genetically different lineages.

Poor growth of the control pistachio UCB-1 trees
The work by Savory et al. indicates that R. fascians and PBTS Rhodococcus 1 and 2 were inoculated on UCB-1 seedlings and no effects were observed (Figure 7-figure supplement 1; panel E). Savory et al. provided no details in regards to the germination, size of seedlings, or growth conditions. In our work, micropropagated, clonal UCB-1 plants were used to mimic the propagation type of affected PBTS trees in the field while minimizing the potential influence of host variability on symptom expression (Stamler et al., 2015a). Based on the data provided in Savory et al. (Figure 7-figure supplement 1; panel E), all of the UCB-1 pistachio seedlings surprisingly exhibited little to no growth after 60 days. In contrast, the mock-inoculated clonal trees used by Stamler et al., 2015a continued to grow rapidly for more than 100 days, and it was nearly 80 days post-inoculation before statistically significant differences in height were observed between inoculated and mock-inoculated trees (Figure 1). The poor growth of mock-inoculated UCB-1 seedlings (Savory et al., 2017) suggests that experimental conditions were not conducive to healthy pistachio growth, thus rendering the control trees unsuitable as a basis of comparison.

Misrepresentation of previous work
i. 'There was a targeted search for Rhodococcus' (referring to Stamler et al., 2015aStamler et al., , 2015b. This is not true, as many potential causative agents were evaluated. ii. 'Bacteria were inexplicably cultured from asymptomatic leaves distal to symptomatic tissues'. If Savory et al. are implying that R. fascians can only be cultured from symptomatic tissues, this has been known for decades (Lacey, 1939;Cornelis et al., 2001). iii. 'Some of the results of the molecular detection were clearly artifacts,' referring to Stamler et al., 2015a. Savory et al. support their statement by stating that the published fasD fragments are identical in sequence, yet they fail to mention that the fasD gene over its entire length is identical among most of the ornamental isolates (Creason et al., 2014). iv. By stating, 'We recognize the insurmountable challenge in showing that there is no possibility that pathogenic Rhodococcus causes pistachio bushy top syndrome, ' Savory et al. admit to the fact that their data did not show a misdiagnosis. We remain confident in our published results.