R/08-05 Molecular Biology of Paralytic Shellfish Poisoning: Dinoflagellate-Bacterial Interactions (completed 1996)

R/08-07 Molecular Biology of Paralytic Shellfish Poisoning: Role of Prokaryotes in Toxin Production (completed 1998)

R/95-01 PSP: Identification and Characterization of Saxitoxin Genes (completed 2001)

R/95-02 Paralytic Shellfish Poisoning: Characterization of Saxitoxin Genes

A metagenomics approach will be the primary method pursued. Briefly, large fragments of DNA from toxic cyanobacteria will be cloned into bacterial artificial chromosomes (i.e., pBAC vectors) and transformed into E. coli. Transformants carrying the "saxitoxin genes," and hence the genetic material required for saxitoxin synthesis, will be initially identified via HPLC and subsequently confirmed by more rigorous assays. Experiments that include cloning of "saxitoxin genes" can be conducted under Biosafety Level 1 (the lowest level) according to NIH guidelines, and hence are not considered to be a significant health risk.

A backup method, amplified fragment length polymorphism (AFLP), is also presented. This method is basically a modified differential display technique. We have identified growth conditions (N2-N as nitrogen source) that promote high levels of saxitoxin accumulation as well as growth conditions (urea-N as nitrogen source) where saxitoxin cannot be detected. We propose to identify differentially expressed mRNA from cells grown under these two growth conditions by AFLP.

Toxic cyanobacterial strains have been identified and grown in the lab for several years. Bacterial contaminants have been greatly reduced in numbers and strains capable of growth on agar plates have been isolated. Growth conditions that result in elevated toxin levels have been identified and, a more difficult task, growth conditions that seemingly result in a complete block of toxin synthesis have also been identified.

September 2002
The past year has been one of change. Our work for the past several years had focused on bacteria as sources of saxitoxins in marine environments. Evidence generated in several labs, including work done by Tracie Baker in our lab, indicates that bacteria do not synthesize saxitoxins. Marine bacteria do, however, synthesize a variety of small compounds that bind to sodium channels (and displace 3H-saxitoxin) or elute from HPLC columns with characteristics similar to saxitoxins. However, detailed chemical analysis of several of these compounds has revealed them to be saxitoxin "impostors."

We are now approaching our goal (i.e., identification of the "saxitoxin genes") from a different perspective. Unambiguous data have confirmed that cyanobacteria synthesize "saxitoxins" and our efforts are now more clearly focused on this group of toxic microorganisms. We are working with two toxic strains, Aphanizomenon flos-aquae and Anabaena circinalis. These cyanobacteria are somewhat difficult to grow, but we have perfected methods for growth of A. flos-aquae and have made significant advances in this area with A. circinalis. These are seemingly small accomplishments, but are absolutely crucial to the success of our program.

Our current objective is to create a pBAC library of A. circinalis in E. coli DH10B. Once generated, E. coli transformants will be screened for toxin synthesis using HPLC methods, with putative positives subjected to more robust screening (e.g., nuclear magnetic resonance, receptor binding assays, etc.). This is a high-risk approach and, as such, it is crucial that graduate students working on this project be assigned sub-projects with a reasonably high probability of success. Toward this goal, Andy Krohn, a new graduate student in the lab, is moving toward generating a genomic map of A. circinalis for his M.S. thesis. His strategy will be to generate a pBAC library, an objective congruent with our overall strategy, and to perform end-sequence analysis of several large inserts. These "end-sequences" will then be aligned with the genomic map of Anabaena sp. PCC 7120, a non-toxic strain whose genome has now been completely sequenced and annotated. Of course, this approach is not without pitfalls. The genome of A. circinalis may have undergone considerable recombination relative to the non-toxic strain, but we have no way of knowing the extent of recombination until the project is under way. In fact, with the exception of E. coli (strains K12 and 0157:H7 the "hamburger strain") the complete genome sequences of two closely related bacteria have not been published. The overall genomes of these two E. coli strains are remarkably similar, and if A. circinalis and PCC 7120 are equally similar, Andy should have no problem generating a map of the saxitoxin-producing cyanobacterium. With luck, Andy may succeed in mapping the relative positions of the "saxitoxin genes" on the A. circinalis genome (i.e., pathogenesis islands were "easily" identified in maps of the toxic 0157:H7 strain of E. coli).

Andy has made progress in generating large DNA fragments suitable for construction of a pBAC library and is in the process of creating a small pUC library for a class project. His overall goal is unchanged (i.e., creating a genomic map) with the caveat that he will temporarily focus on pUC for creation of a genomic partial library. This is a valid short-term approach as the pUC vector system is 10-100X easier than the pBeloBAC vector system. Andy is focusing on relatively large fragments (i.e., 3-4 kB) in anticipation that fragments of this size are likely to carry two genes. End sequence analysis of these fragments will provide Andy with the first, and far easier, idea of the extent of recombination in A. circinalis relative to PCC 7120. Andy is currently taking a bioinformatics course and will potentially bring exciting new computational methods into our lab that will be helpful as this overall project proceeds.

Krohn, Andrew. M.S. Chemistry/Biochemistry, University of Alaska Fairbanks.

Matweyou, Julie. M.S. Biological oceanography, University of Alaska Fairbanks. "The relationship between Alexandrium abundance and paralytic shellfish toxins on Kodiak Island, Alaska."

Finally, our earlier work with symbiotic bacteria associated with toxic strains of Alexandrium (Plumley et al. 1999. J. Phycol. 35:1390-96) caught the eye of a Canadian Ph.D. student. Andrew Lang subsequently applied for a Natural Sciences and Engineering Research Council of Canada (NSERC) Post-doctoral Fellowship to work in our lab and recently learned that his application was successful. Andrew Lang will pursue a Tn5 mutagenesis approach to identify bacterial genes involved in saxitoxin production in symbiotic relationships with Alexandrium. An immediate goal will be to characterize the histidine kinase sensor protein we previously identified and to determine if it plays a role in initiation or maintenance of the symbiotic relationship between Pseudomonas stutzeri and Alexandrium lusitanicum. Andrew Lang is scheduled to arrive in February 2003.

Andy Krohn's undergraduate project was entitled Molecular Techniques Involved in PCR Amplification and Cloning.

Submitted Publication
Tracie R. Baker, Gregory J. Doucette, Christine L. Powell, Gregory L. Boyer and F. Gerald Plumley. GTX4 Imposters: Characterization of Fluorescent Compounds Synthesized by Pseudomonas stutzeri SF/PS and Pseudomonas/Alteromonas PTB-1, Symbionts of Saxitoxin-Producing Alexandrium spp. Toxicon. (accepted pending minor revisions).