CICEET Progress Report for the period 9/02/07 Through 3/01/08

Project Title: High-throughput Quantitative Detection of Microbial Contaminants
Principal Investigator(s): Mara R. Diaz
Additional Investigator(s): Kelly Goodwin and Jack W. Fell
Project Start Date: September 2007 through March 2008

Figures

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

II. Lineal RCA:
In view of the fact that Mirus Labeling IT technique showed more potential for our application (see section G), we followed the suggestion of our advisor and concentrated all our efforts toward the development of Mirus label IT technique. Mirus label IT technique is a non-enzymatic direct chemical labeling system that allows the covalent attachment of biotin molecules to guanine and other active heteroatom residues of DNA.

Tasks to meet objectives
i) Two different labeling mechanisms were compared and assessed with D1/D2 amplicons derived from large subunit (LSU) rDNA gene. Chemical labeling used Mirus Label IT, a non-enzymatic direct chemical labeling system that allows the covalent attachment of biotin molecules to active heteroatoms in the nucleic acid. The enzymatic labeled method used single biotinylated amplicons generated by PCR and 5’ end biotinylated reverse primer. The performance of both labeled targets was established using a direct hybridization assay.

ii) The hybridization specificity of Mirus labeled target was assessed with D1/D2 amplicons representing closely and non-closely related dinoflagellate species. The experiments were undertaken at two different incubation temperatures.

iii) Genomic DNA labeling efficiency was determined with different ratios of Label IT reagent to nucleic acid. In addition, the incorporation of biotin moieties was indirectly measured by spectrophotometry.

iv) Optimization of hybridization conditions employed digested and non-digested WGA (whole genomic isothermal amplification) DNA and involved different incubation times and temperature conditions. Double digestion reactions were undertaken with WGA products of K. mikimotoi and K. brevis. The reaction used the enzyme cutters AIW6 and MBOII. The digested products were assessed by electrophoresis.

Progress on Tasks
Progress on tasks is described in section G (Preliminary data section).

Have the results/data gathered during this reporting period changed the project objectives when compared to your original proposal? Please explain.
The main objective of the project has remained the same.

Dissemination activities during this reporting period (please include the number of participants where applicable).
Preliminary data was recently presented at UM OHH Center Annual Scientific Symposium in Miami, FL. Jan 2008

Difficulties
The availability of K. mikimotoi cultures and/or genomic DNA continues to be a limiting factor. Therefore, as previously stated, some reactions used amplicons derived from whole genomic isothermal amplification. Amplified genomic DNA was diluted and subsequently used for our downstream application. Another difficulty is that some probes, which showed an excellent specificity with amplicon targets, did not meet the specificity requirements or failed to hybridize when tested with genomic DNA or whole genomic amplicons (WGA). The specificity problem, might be associated with secondary binding sites and/ contaminants in the DNA.

Data Generated to date
i)Comparison of two labeling methods: chemical (Mirus Label IT) versus enzymatic labeling (end labeling with a biotinylated primer). In order to evaluate the labeling efficiency and the feasibility of Mirus Label IT for our downstream application, we compared Mirus Label IT technology with the enzymatic 5’terminal labeling method. The experiment, which was carried out with PCR amplicons, showed that the chemical Label IT system outperformed the enzymatic 5’terminal labeling technique (Figure 1). All three probes (i.e Pkb, Pkb2 and Pkmiki) showed significantly higher signal fluorescent intensities when challenged with amplicons labeled with Mirus Label IT technique (Figure 1). The enhancement in signal intensity ranged from 75% to 80%. These results illustrate that Mirus Label IT, as opposed to 5’end labeling (which allows one label per strand), appears to be a more effective labeling system since multiple biotin moieties can be incorporated on any reactive heteroatom in the polynucleotide. Although, N7 of guanine was previously reported as the preferred heteroatom, new evidence indicated that other heteroatoms such as the N3 of adenine and the N3 of cytosine react as well. Based on this data, we are very confident that both amplification technologies (target and signal amplification) can be merged to create a powerful method for the detection and occurrence of HABs species at pre-blooming conditions.

ii) Effect of Mirus labeling on hybridization specificity. To evaluate if Mirus Label IT can affect the assay specificity, Label IT amplicons representing closely and non-closely related dinoflagellate species were tested. The assays were undertaken at 50°C and at 30min (data non shown) and 1.5hrs incubation time intervals (Figure 2). The results showed that Pkb and Pkmiki were highly specific to their complementary targets. No cross-reaction was found among any of the tested species at any of the incubation time intervals.

iii) Determine optimum labeling efficiency with genomic DNA. The labeling efficiency was assessed with 1µg of K. mikimotoi WGA DNA and 1 to 5 µl of Label IT reagent. After three hours incubation at 37°C, 167 ng of DNA was hybridized with Pkmiki (Figure 3). A subtle increase in MFI signal occurred as the amount of labeling reagent was increased from 1 to 4 µl. However, a significant reduction in signal was observed when the reaction employed 5 µl of the dye. At those levels, over saturation of the Label IT reagent might explain the observed “hook effect”.

Incorporation of the labeling material was also assessed with a NanoDrop ® ND-1000 spectrophotometer. For this particular assay, the labeling density was determined indirectly and used 1-4 µl of Cy3 Label IT. DNA molar concentration was determined using absorbance at 260 nm and the calculated molecular weight and extinction coefficient for each DNA oligonucleotide. Although the label estimation used an indirect approach, this method gives fair estimates of the extent of the label density (B. Shannon, personal communication). The results are described below:

1µl of labeling =50pmol/µg ~ 30base pairs of double-stranded DNA per label
2µl of labeling =100pmol/µg ~ 15base pairs of double-stranded DNA per label
4µl of labeling =150pmol/µg ~ 10base pairs of double-stranded DNA per label

Based on the hybridization and the spectrophotometric analysis, we selected a labeling ratio of 1:2 (1 ug of DNA: 2ul of Label IT reagent) as this level of labeling is sufficient to allow an adequate yield of labeling.

v) Optimization of hybridization conditions with genomic DNA.
a) Effect of temperature on hybridization. Hybridization assay conditions were conducted at various temperatures (35°C and 50°C) (Figure 4). Pkmiki showed a gradual decrease in signal intensity as the temperature of the assay was increased, with the lowest signals recorded at 50°C. Conversely, Pkb failed to hybridize with its target WGA genomic DNA at any of the chosen conditions (data not shown). In contrast, robust fluorescence signals of up to ~3000 MFI were obtained when Pkb was challenged with its target PCR amplicon (data not shown). Secondary structures at the binding site of WGA DNA might be impeding the accession of Pkb to the target site. In view of this unexpected outcome, two additional probes, Pkb2 and Pkb658, were designed and successfully validated with their target amplicons (data not shown).

b) Time course incubation experiments. In order to determine the optimum incubation time, the assay was monitored at different time intervals ie., 2 hrs, 3hrs, 4hrs and 5hrs. The assay was carried out at 45°C and used 167 ng/well of K. mikimotoi WGA DNA. The data showed a proportional increase in fluorescence signal intensity as the time of incubation increased (Figure 5). However, no substantial increase in hybridization signal was obtained after 4-hrs. Also, no cross reactivity was found when K. mikimotoi WGA DNA was challenged with Pkb.

c) Evaluation of restriction enzymes with genomic DNA. Double digest reactions were undertaken with the enzyme cutters, AIW6 and MBOII. The digested WGA DNA products were assessed by electrophoresis and tested with our direct hybridization assay format. The assay used two different temperature conditions and 4 hrs of incubation. The hybridization of undigested DNA consistently showed higher hybridization efficiency than double digest products (Figure 6). For instance, at 35°C and 45°C the undigested K. mikimotoi DNA displayed signal intensities ranging from 1,280 to 715 MFI, respectively. In contrast, double digest products displayed fluorescence signals ranging from 830 to 250 MFI. Although it is expected that shorter fragments hybridize more efficiently than longer fragments, the observed difference in signal intensities could also be due to the different number of fluorescent molecules that both targets may carry. In theory, native or unfragmented DNA should carry more fluorescence molecules, which can account for higher fluorescence signals. Interestingly, similar results were observed when the assay was conducted with digested and non-digested amplicons (data not shown). The data also show that at 35°C the assay does not provide a good level of stringency since the species K. mikimotoi cross-reacted with Pkb. However, when the assay was conducted at 45°C, the cross-reactivity was abolished (Figure 6).

d) Probe specificity at various temperatures and time intervals. To determine the assay specificity, Pkmiki probe was challenged with the closely related species K. brevis. K. brevis differs from the probe sequence by a single bp at the 5’end and two centrally located mismatches. In contrast to the excellent level of specificity, which was attained with PCR amplicon targets (Figure 2), the species K. brevis (unfragmented WGA DNA) displayed some cross reactivity with the probe Pkmiki. As illustrated in Figure 7, a slight increase in cross reactivity was observed as the incubation time increased. At 4 hrs incubation the cross reactivity levels reached ~ 130 MFI, which accounted for ~19% of specific hybridization signal.

Locked Nucleic Acids
Although not included in our original work plan, and in order to circumvent some of the observed specificity or affinity issues, we modified various probe sequences by incorporating a few locked nucleic acid bases (LNAs). The LNA bases contain a methylene bridge that connects the 2’oxygen of the ribose with the 4’ carbon of the ribose ring. The presence of this bridge locks the sugar ring in one fixed conformation, thus allowing a decrease in the entropic penalty of hybridization. This in turn, increases the affinity and specificity of the probe toward its target (Beane et al., 2007). In addition, duplexes with LNA residues display an increase in melting temperatures by approximately 2-4°C per LNA modification.

Two 19mer probe sequences, Pkmikilna and Pkb2lna, were designed with three consecutive LNA bases (green font). The modifications were located around positions where specificity and discrimination was needed. See below.

K. brevis Pkb2lna Tm: 61°C
TGCTTCCGGGCGACTGAAT K. brevis
cGCTTCCGaGtGACTGAAT K. miki

K. mikimotoi Pkmikilna Tm: 57°C
CGCTTCCGAGTGACTGAAT K. miki
tGCTTCCGgGcGACTGAAT K. brevis

In order to evaluate the effect of LNA modification on probe signal, unmodified (Pkb2 & Pkmiki) and modified (Pkb2lna & Pkmikilna) versions of the probe sequences were initially tested with 5’end labeling and Mirus Label IT amplicons (Figure 8 and Figure 9). The results showed that incorporation of LNA residues to probe sequences enhanced the fluorescence signal intensity of Pkmikilna and Pkb2lna by 11.5 and 33%, respectively (Figure 8). The latter data was generated with biotinylated PCR amplicons. In contrast, nearly identical signal intensities were documented with Mirus Label IT amplicons (Figure 9).

To determine the performance of LNA probes with genomic WGA DNA and to determine whether the optimal hybridization temperature condition is applicable for LNA probes, each probe was tested with their respective WGA target for 4 hrs at 45, 50, and 55°C (Figure 10). The data showed a significant decrease in hybridization efficiency when the assay was conducted at 55°C, but a small decrease in fluorescence intensity was documented at 45°C and 50°C. As opposed to previous results, modifications of Pkmiki with LNA bases allowed successful hybridization with WGA target at 50°C. Contrary to Pkb, which failed to hybridize with its WGA target, the newly designed Pkb2lna, successfully hybridized with K. brevis WGA target (Figure 10).

To optimize assay conditions, LNA probes were tested at different temperatures ranging from 50 to 57°C. The hybridization assay was carried out for 4hrs and employed 167ng of unfragmented native genomic DNA. As expected, Pkb2lna showed a gradual decrease in hybridization efficiency as the temperature of the assay was increased (Figure 11). However, despite high temperatures, robust signals were documented at 57°C. This data is in sharp contrast to previous experiments undertaken with WGA DNA (Figure 10), which showed significant lower MFI values and an overall reduction in hybridization efficiency at 55°C. Although, the data showed that Pkb2lna has an excellent affinity with unfragmented native genomic DNA, an opposite scenario was documented with Pkmikilna since this probe basically failed to hybridize with unfragmented or sheared genomic DNA target (data not shown). In view that this particular experiment employed a different strain (K. mikimotoi ccmp 430), we corroborated the identity of the strain using our developed PCR-Luminex array platform. The assay confirmed the identity of ccmp430 as K. mikimotoi (data not shown). In this scenario, it can be argued that hybridization failure was due to structural conformations around the probe binding site that is not permitting K. mikimotoi unfragmented native genomic DNA to bind to its capture probe, and that shearing treatment (15 passes with 1 ml syringe-25 gauge) did not suffice to eliminate potential secondary structures. We contemplate to either increase the number of passes or shear stress the DNA with an ultra-sonicator.

In order to determine the specificity of the LNA probes, the assay was undertaken with Mirus label IT amplicons representing various dinoflagellate species. As shown in Figure 12 , none of the dinoflagellate species significantly cross reacted with LNA probes, confirming the good affinity of the probes with their respective amplicon targets. A similar experiment was undertaken with unfragmented native genomic DNA (Figure 13). In this case, we tested the performance of Pkb2lna with unfragmented native genomic DNA. The experiment, which initially was carried out at 56°C for 4hrs, showed that some of the tested species, i.e. K. mikimotoi (two strains), Gonyaulax cocleapoly and Pyrrophyta spp. strongly cross-reacted with Pkb2lna. However, when the hybridization time was reduced to 2.5 hrs, the cross reactivity levels decreased by ~40 to 75% (Figure 13). Despite this significant signal reduction, the cross reactivity displayed by ccmp430 and G. cocleapoly was still quite significant. In view of this outcome, D1/D2 species amplicons were challenged with K. brevis probes: i.e., Pkb, Pkb2, Pkb2lna and Pkb658. This step was adopted to screen samples for potential cross-contamination. Interestingly, from all the species tested, G. cocleapoly displayed a very low cross-reactivity (~100 MFI) when challenged with Pkb2 and Pkb2lna. Although these MFI values are quite low, they could be indicative of a possible cross-contamination with K. brevis DNA. Other possible explanations for cross-hybridization include: a) presence of a secondary binding site outside our target LSU region; b) sequence variation within the copies of rDNA locus. It has been estimated that dinoflagellates contain approx. 200 copies of rDNA gene and sequence variants among these copies can contribute to cross-reactivity; c) presence of a prokaryotic symbiont.

Dinoflagellates are known to harbor a bacterial flora that can be attached to algal cell walls or within algal cells. These complex flora are often present in many dinoflagellate cultures since they are difficult to eliminate due to the time consuming procedures that involve washing with sterile diluents, serial dilution, ultrasonic treatment, differential centrifugation and antibiotic treatments. Although it is feasible to generate bacteria-free culture, re-incidence of bacteria in axenic cultures is not uncommon. Therefore, the observed cross-hybridization could have resulted from a symbiont or free-living bacteria, which has sequence similarity with Pkb2lna. A BLAST sequence similarity search in GenBank retrieved several sequences that differed by only 2 bp from our probe. These recent submissions were classified as marine metagenomes. These metagenomes represent a pooled of bacterial genomic DNA recovered from environmental samples. If metagenomes are indeed responsible for the cross hybridization, then they could explain the disparity pattern documented when the hybridization assay was performed with native genomic DNA and D1/D2 amplicons. In view that bacteria were unable to amplify with our primer sets and PCR conditions, as they do lack the D1/D2 region, they were not expected to compete for amplification and hence did not have the likelihood to induce cross hybridization. Otherwise, similar cross-reactivity patterns would have been documented for both type of DNA templates. In order to elucidate the possible presence of bacterial DNA, dinoflagellate genomic DNA will be amplified with 16S-23S ribosomal primers (Prokic et al. 1998). Any positive PCR product band will be indicative of the concomitant presence of bacteria in the sample. The PCR product will be further screened with our Pkb2lna probe.

Project Objectives for Next Reporting Period

Objectives
Mirus labeling i) Evaluation of other restriction enzymes or sonication with native genomic DNA
ii) Re-design of Pkmiklna
iii)Amplification of bacterial DNA symbionts to evaluate possible presence of bacterial DNA
iv) Re-design of Pkb2lna or Pkb658
v) Optimization of hybridization conditions with redesigned probes or potential new probes
vi)Test specificity of re-designed probes
vii)Determine detection levels of the prospectus probes with native genomic DNA
viii)Test prospectus probes with environmental samples

Work Plan to Meet Objectives
i) In view that Karenia genome is rather complex, the selection of a proper enzyme cutter is necessary to reduce to some extent the bulky nature of its genome. Therefore, we will further undertake the testing of other restriction enzymes or will employ DNA sonication.

ii) Since the performance of Pmiki probe was rather poor at hybridization temperatures higher than 50ºC, we will incorporate more LNA residues to enhance the melting temperature of the probe. The residues will be located at strategic position to boost the specificity of the probe. This strategy will allow for more stringent conditions.

iii) In order to elucidate the possible presence of bacterial DNA, the dinoflagellate genomic DNA samples will be amplified with 16S-23S ribosomal primers (Prokic et al. 1998). Any positive PCR product band will be indicative of the concomitant presence of bacteria in the sample. The PCR product will be further screened with our Pkb2lna probe.

iv) Pkb2lna probe will be redesigned to abolish potential cross-hybridization with marine metagenomes. For this purpose, we will lengthen the probe sequence by few bp and will introduce more LNA residues at key positions. In addition, we will modify Pkb658 by adding some LNA residues to the probe sequence.

v) Optimization of hybridization conditions with redesigned probes. This will involve different incubation times and temperature conditions. The optimum conditions will be determined by the median fluorescence intensity values.

vi)Specificity of the assay will be assessed with closely related and non- closely related species of K. brevis and K. mikimotoi.

vii) Detection levels will be assessed with native genomic DNA and will be established with serial dilutions.

viii) Collect environmental samples pre-, during, and post- K. brevis bloom

Dissemination Objectives for next reporting period
We expect to present our results in American Society for Limnology and Oceanography.

Overall Project Timeline Update
We will allocate the time to accomplish the tasks described in section 2.

Expenditures
The money allocated for the expenditures seems appropriate

End User Advisor Feedback
Advisor: Dr. James Jacobson
Organization: Luminex Co.
Location: Austin, Texas
Phone number: 512-219-8020
E-mail: jwjacobs@luminexcorp.com

At this stage, what are the potential applications for this research? Please discuss how you and others could potentially use the technology.
The progress that has been demonstrated in this report with sample preparation and labeling methods has moved the project forward nicely. Providing solutions for these assay application challenges means that a rapid, simplified and geographically distributed test for these targets is within reach. As was discussed in the original project proposal, previous interim reports and in the current report, harmful algal bloom (HAB) continues to have significant economic and health-related impacts. Methods developed for detection of HAB agents can readily be extended to include other environmental monitoring situations where pathogens, contaminants, toxins, etc. decrease the economic value and / or safety to those using the beaches, fields, parks or similar areas. However, the most common challenge in all of these scenarios is first a sampling problem ­ how do you develop a method sensitive enough to detect the minimum number(s) of targets in the test volume or quantity of material? And, secondly, how can one develop a simplified, robust assay that meets the sensitivity and specificity requirements and is still practical for a wide range of users and locations. Results detailed in this report show significant progress in both of these areas.

What are the key challenges to application of this technology? Please consider the technology itself as well as issues related to regulation, politics, socio-economic pressures, trends in the field etc.

The work summarized in this progress report has focused on identifying and optimizing a target labeling method that is both sensitive and unbiased in broad application. In my past communications, I have encouraged the investigators to focus on this challenge as an immediate goal. I am pleased with the progress that has been made. The project objectives and work plan outlined for the next reporting period are logical and necessary extensions of this work. Once the label / detection protocol is in place, the other assay requirements ­ sample type, sample volume, incubations, workflow, etc. ­ will allow commercialization and distribution of the application.

Has anything changed about this project's potential applicability since the last reporting period (not applicable to the first Progress Report)?
There continues to be an urgent need for this application. The progress detailed in this interim report is very encouraging and gives increased confidence that the approach will succeed.

Questions/comments/ suggestions for the researchers?
Nice work ­ keep it up. In addition, always keep the end user in mind and keep it simple!

PI Response to End User Advisor Feedback
Nothing to reply.

Bibliography
Beane, R., Ram, R., Gabillet, S., Arar, K., Monia, B., and Corey D. R. (2007) “Inhibiting gene expression with locked nucleic acids (LNAs) that target chromosomalDNA. Biochemistry 25:7572-7580.

Prokic, I, Brummer, F., Brigge, T., Gortz, H.H., Gerdts, G., Schutt, C., Elbrachter, M., Muller, E.G. (1998) "Bacteria of the genus Roseobacter associated with the toxic dinoflagellate Prorocentrum lima." Protist 149: 347-357.