Embed Size (px)
Citation preview
Cyanotoxin Analysis Methods
Oregon Cyanotoxin Rule Considerations Webinar August 22, 2018
Heather RaymondOhio EPA HAB Coordinator
Analytical Method Comparison & Microcystin Variant Evaluation
• 11 Sites: 4 up‐ground reservoirs, 2 in‐stream reservoirs, 2 Lake Erie locations, 2 canal‐feeder lakes, and 1 river source
• 22 samples from 2014 selected to help evaluate spatial and temporal variability within source waters
• Variety of cyanobacteria genera represented• Each sample analyzed using 5 separate analytical methods
Methods Evaluated
• Enzyme‐Linked ImmunoSorbent Assay (ELISA) Microcystin‐ADDA Method
• Liquid Chromatography (LC) – Ultraviolet (UV)• Liquid Chromatography(LC) –Tandem Mass Spectrometry (MS/MS) (individual variants)
• LC‐MS/MS (MMPB)• LC‐UV and LC‐MS (scan for variants without standards)
Microcystins Testing
- Over 300 microcystin variants - Standards not available for majority of variants
No “Perfect” Analytical Method for Detecting TOTAL Microcystins
ELISA Microcystins‐ADDAEnzyme‐Linked ImmunoSorbent Assay (ELISA) Microcystin‐ADDA Method (detection of antigen using an antibody)
– Measures total microcystins (all variants/congeners, based on ADDA)– Highly selective/specific (for ADDA)– Certified by ETV Program– Moderately sensitive (RL: 0.30ug/L)– Suitable for raw & finished water – Quick (~four hours), useful for operational adjustments– Relatively inexpensive– Does not require high end equipment or expertise to run (can
(can be used in water system lab)– Does not require pre‐concentration solid phase extraction (SPE) step – Does not provide concentrations of specific microcystin variants– Is an indirect measure of the toxin– Non‐linear curve: may require sample dilution and reanalysis if results out of
range
– Ohio EPA Standard Method 701.0 & Lab Certification– U.S. EPA Method 546
Liquid Chromatography (LC) – Ultraviolet (UV)
LC‐UV‐ Liquid Chromatography separates components‐ Microcystins have UV absorption maxima at 238 nm‐ Non‐selective detector; co‐eluting interferents prevent accurate identification of components and quantitation
‐ Less expensive than mass spectrometry ‐ Less sensitive than mass spectrometry (average LOQ ~ 0.3 µg/L)
‐ ISO 20179 Standard Method
Liquid Chromatography(LC) –Tandem Mass Spectrometry (MS/MS)
• LC/MS/MS– Highly specific identification of components (based on standards)
– MS can identify a component in the presence of co‐eluting interferentsbut quantitation may be compromised
• Presence of co‐eluting interferents can act to suppress or enhance response resulting in analytical bias
• Sensitive (LOQ ~ 0.02 µg/L)– “Weak” product ion abundance limits sensitivity. Requires pre‐
concentration with SPE to augment sensitivity (LOQs < 0.02 µg/L)• Preconcentrates NOM too
– U.S. EPA Method 544• Standard Method‐ includes QA/QC protocols and reduces variability in results
between labs• Limited to 6 microcystin variants and finished water only
– Expensive and requires highly skilled analysts
– Issues with standard availability, purity, and variability
Use of Standard Addition to Account for Matrix Effects in LC‐MS/MS Analysis
LC‐MS/MS MMPB Method– MMPB (2‐methyl‐3(methoxy)‐4‐phenylbutyic acid) method analyzes the
chemically cleaved Adda group common to all microcystin variants – Measures total microcystins (all variants, based on ADDA)– Quick (~2 hours, does not require freeze/thaw or sonication)– Sensitive (0.05 ug/L)– Does not require standards for individual variants– Utilizes 4 PB internal standard– Suitable for raw water, some limitations with finished water– Does not provide data on individual variants– Requires oxidation step– Potential for detection of microcystins disinfection byproducts
Toxicon 104 (2015) 91-101 (Foss & Aubel): Using the MMPB technique to confirm microcystin concentrations in water measured by ELISA and HPLC (UV, MS, MS/MS)
LC‐UV/PDA & LC‐MS ScanUses two LC‐based methods in tandem to independently confirm presence of microcystins
– Can detect microcystin variants without standards– No standard methods, expensive, requires complex data‐interpretation,
time‐consumingSource: Greenwater Labs
* LC-UV data presented does not include false-positives that were eliminated from total (Based on lack of confirmation with LC-MS methods). Sample # 14 was non-detect using LC-UV.
Results of Method Comparison
Ohio-EPA-10x-166165-E-Fork-Camp-Beach... 3/16/2015 4:38:28 PMKinetexC18
RT: 3.98 - 20.00 SM: 15G
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Time (min)
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e Ab
unda
nce
7.27
4.84
4.767.53
7.09
4.40 8.51 14.28 14.5913.2912.8712.41 16.6911.86 16.118.36 17.45 17.74 19.286.19 15.38 18.9510.239.80 10.965.15 5.71 6.71 9.41
NL:
5.49E2
TIC MS
Ohio-EPA-10x-
166165-E-Fork-
Camp-Beach-
MC-MSMS-
031615-2
RT: 4.23 - 19.56
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Time (min)
0
2000
4000
6000
8000
10000
12000
14000
16000
uAU
16.65
17.13
16.14
17.81
15.4718.27
15.048.144.40 7.275.06 8.797.517.14
4.78 9.709.276.56 14.336.08 14.035.76 10.06 10.74 11.17 13.6111.70 12.11
NL:
1.73E4
Channel A UV
Ohio-EPA-10x-
166165-E-Fork-
Camp-Beach-MC-
MSMS-031615-2
UV chromatogram with multiple peaks, most not corresponding to MCs (SPE was used)
MS/MS TIC & individual variant chromatograms
Analysis: Amanda Foss, Greenwater Labs
Results of LC‐MS/MS MMPB and Individual Variant Analysis Compared to ELISA
Inland Lake Microcystin Variants (Planktothrix)MC‐Variant Site 1
6/16/14Site 26/16/14
Site 29/2/14
[DAsp3] MC‐RR 5.3 6.1 17.5[Dha7] MC‐LR 1.1 1.4 1.5MC‐YR 0.2‐0.6 0.2‐0.6 1.2MC‐RR 0.1‐0.3
Inland Lake Microcystin Variants (Mixed Bloom)MC‐Variant Site 1
6/18/14Site 26/18/14
Site 27/9/14
Site 36/30/14
[Dha7] MC‐RR 2.9 3‐9 1.0 0.08MC‐RR 1.4 39 1.0 0.01‐0.03MC‐YR 1.1 15 1.0MC‐LR 4.0 67 2.4 0.55[DAsp3] MC‐LR 0.6 18 0.4 0.03[Dha7] MC‐LR 3.6 1.0 0.05MC‐WR 0.2‐0.6 0.2‐0.6MC‐LA 0.2‐0.6MC‐LY 0.2‐0.6 6 0.2‐0.6 0.10
MC‐Variant Lake Erie UpgroundReservoir
In‐streamReservoir
CanalLake
Stream
MC‐YR Y Y Y Y[Dha7] MC‐LR Y Y Y Y[DAsp3] MC‐RR Y Y Y YMC‐LR Y Y YMC‐RR Y Y Y YMC‐LY Y Y Y YMC‐WR Y Y Y[DAsp3] MC‐LR Y YMC‐HilR YMC‐LA Y Y[Dha7] MC‐RR Y YMC‐FR Y[DAsp3] MC‐FR Y6.9 min 1049.5 m/z Y7.5 min 1029.5 m/z Y8.7 min 1043.5 m/z Y
Microcystin (MC) Variant Distribution by Source Type
MC‐LF and Nodularin, which are included in USEPA Method 544, were not detected (MC‐LF and additional MC variants have been detected in follow‐up studies).
Key Findings• 16 different MC‐variants were detected • MC‐LR was only detected at 5 of 11 sites (45%)• Most common variants were: MC‐YR, [Dha7] MC‐LR and
[DAsp3] MC‐RR• LC‐PDA methods prone to interference, potential for false
positives and false negatives• LC‐MS/MS MMPB method helped confirm ELISA results• 91% of samples had MC‐variants not detectable by U.S. EPA
Method 544 (including dominant MC‐variant in some samples)• LC‐MS/MS individual variant analysis under‐reported total
microcystins, based on MMPB and LC‐UV/MS scan data
ELISA MC‐ADDA Matrix Interference StudiesTreatment Chemical Microcystins – ADDA ELISA Assay Tolerance (< / = )
Sodium Carbonate (Soda Ash) ≤25 gpg
Sodium Hexametaphosphate ≤250 ppm
Sodium Silicofluoride ≤10 ppm
Aluminum Sulfate1 ≤100 gpg (with pH adjustment within assay tolerance)
Calcium Oxide (Lime)1 ≤2000 gpg (with pH adjustment to within assay tolerance)
Potassium Permanganate2 ≤10 ppm (with quenching using 1 mg sodium thiosulfate per 1 ml sample)
Sodium Chlorite2 ≤10 ppm (with quenching using 1 mg sodium thiosulfate per 1 ml sample)
Carbon3 ≤2 ppm with filtering at time of sampling1Natural pH of solution outside assay tolerance, tolerance levels determined after pH adjustment2 Oxidizers degrade microcystins, tolerance determined after quenching3 Tolerance level due to effect of carbon on toxin, not assay performance
Lisa Kamp, et. at, 2016. The effects of water sample treatment, preparation, and storage prior to cyanotoxin analysis for cylindrospermopsin, microcystin and Saxitoxin. Chemico‐Biological Interactions.
ELISA MC‐ADDA Matrix Interference Studies
Studies by U.S. EPA as part of ELISA MC‐ADDA Method Development for UCMR 4:• Storage Stability – Holding Times• Sample Preservation and Container Studies• Matrix Interference Studies
‐Microcystins Variant Fortified Sample Studies (finished water, raw water, reagent water with chemical addition, etc.)‐Dilution Experiments (real world raw/finished water samples)
• U.S. EPA Method Validation & Interlab Validation
LC‐MS/MS MMPB Method Finished Water Matrix Evaluation:• Concern regarding detection of microcystins disinfection byproducts
Analytical Methods Utilized by Ohio EPA
Microcystins(μg/L)
Cylindro‐spermopsin
(μg/L)
Saxitoxins(μg/L)
Anatoxin‐a(μg/L)
Surveillance samplingELISA
(MC‐ADDA)ELISA ELISA LC‐MS/MS
Repeat sampling in response to a finished water detection
ELISA(MC‐ADDA)
LC‐MS/MS LC‐MS/MS LC‐MS/MS
ELISA: Enzyme‐Linked Immunosorbent Assay
LC‐MS/MS: Liquid Chromatography followed by tandem Mass Spectrometry
Cyanobacteria Screening: Molecular Methods (Multiplex qPCR)
• Multiplex quantitative polymerase chain reaction (qPCR) assay –identifies and quantifies the presence of genes unique to:
• Cyanobacteria: 16S rDNA• Microcystins and Nodularin production: mcyE gene• Cylindrospermopsin production: cyrA gene• Saxitoxins production: sxtA gene
– Fast: 2‐3 hours– Scalable– Cost‐effective– Utilizes certified reference material – Specific
• Ohio EPA DES 705.0 Method and lab certification• www.phytoxigene.com (current assay in use)• Ohio EPA method evaluation study
Microcystins detected in raw water at 45 PWSs (38%) and mcyEdetected at 57 PWSs (48%)
Out of 1850 paired PWS samples: 100% of microcystins >1.6 µg/L had paired mcyE gene detections. 100% of microcystins >5 µg/L had mcyE detections > 5gc/µL 90% of microcystins detections >1.6 ug/L had mcyE detections
>5 gc/µL Less than 2% of samples (22 sites, 32 samples) had microcystins
detections without mcyE detections: 19 of the 22 sites had gene detections in either prior or post sampling events
The remaining three PWSs had only one low level (0.35 – 0.44 µg/L) microcystins detection in 2016; all had trace mcyE gene copies
qPCR as Screening Tool for Microcystins
Example: Source Water With Consistently High Microcystins & mcyE Concentrations
Microcystins, m
cyE
1
10
100
1000
10000
6/1/16 7/11/16 8/20/16 9/29/16 11/8/16 12/18/16
mcyE gene (gc/µL) Microcystins (µg/L)
Example: Phytoplankton Enumeration versus qPCR
0
500
1000
1500
2000
2500
0
0.5
1
1.5
2
2.5
3
5/30/16 7/19/16 9/7/16 10/27/16 12/16/16
mcyE(GC/µL)
Microcystins(µg/L)
Total Cyanobacteria qPCR (GC/µL)
Total Cyanobacteriacount (cells/µL)
ND ND
Microcystins, m
cyE
Total Cyanobacteria, 16S
0
5000
10000
15000
20000
25000
0123456789
10
Jun‐16 Jul‐16 Jul‐16 Aug‐16 Sep‐16 Oct‐16 Nov‐16 Dec‐16
Microcystins, m
cyE
Total Cyanobacteria, 16S
Microcystins(µg/L)
mcyE(GC/µL)
16S(GC/µL)
Example: 16s and mcyE trends
• Microcystins trend with mcyE genes • Non‐toxic cyanobacteria bloom (16s) in August
ND ND
sxtA detections triggered response sampling at 33 PWSs (22%), including many with no historic saxitoxinsoccurrence and three Lake Erie intakes. 15 of those PWSs detected saxitoxins in raw water (12%) 6 PWSs had finished water detections (none above thresholds) Less than 1% percent of samples had saxitoxins detections
without corresponding gene detections (includes 168 paired inland lakes samples)
Only one cyrA detection, no cylindrospermopsindetections in 2016
cyrA detections at 2 PWSs and cylindrospermopsindetections at one of those PWSs in 2017
qPCR as Screening Tool for Saxitoxinsand Cylindrospermopsin
0.01
0.1
1
10
100
1000
10000
100000
0
5
10
15
20
25
30
5/30/16 7/19/16 9/7/16 10/27/16 12/16/16
Microcystins (µg/L) mcyE (GC/µL) 16S (GC/µL)
Saxitoxins (µg/L) sxtA (GC/µL)
Microcystins, m
cyE
Saxitoxins, sxtA, 16S
ND
Example: Simultaneous Saxitoxins and Microcystins
0.01
0.1
1
10
100
0
50
100
150
200
250
300
350
Cylindrospermopsis PlanktothrixAphanizomenon Dolichospermum (Anabaena)Saxitoxins sxtA
Site A B C D1 D2 D3 D4 E1 E2 F
Cyan
obacteria
(cells/
µL)
Saxitoxins(µg/L) &
sxtA(GC/µL)
Inland Lake Examples: Saxitoxin‐producing Cyanobacteria, sxtA, and Saxitoxins
• Low concentrations of saxitoxins detected in drinking water (below Ohio EPA threshold) from late July – September, 2015.
• Extracellular saxitoxins predominated all samples.• 10 different potential saxitoxin‐producing genera found in multiple
habitat zones (pelagic, benthic, periphyton, etc.) in multiple (all sampling) locations.
• Follow‐up qPCR analysis revealed highest sxtA gene copies in benthic samples.
• qPCR data used to target algaecide application.
Example: Use of qPCR to Help Identify Source of Saxitoxins and Target Reservoir Management
qPCR Summary mcyE is an effective, specific, screen for microcystins
and may be useful as an early warning tool. sxtA is an effective, specific, screen for saxitoxins. Multi‐plex functionality of assay works. 16S is not an effective screen for cyanotoxins, but
potentially useful for assessing susceptibility qPCR is superior to phytoplankton enumeration as a
screen for saxitoxins and provides a more targeted screen for microcystins.
qPCR data can help inform reservoir management strategies.
Adapted from:Detection and Monitoring of
Microcystins in Theory & Practice
Sheela Agrawal, Ph.D., NEORSDJudy Westrick, Ph.D., Wayne State
University
NEORSD vs WSU
y = 1.1038x + 1.0934R² = 0.831
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18
Total M
C µg
/L by WSU
Total MC µg/L NEORSD
Total MC Data WSU vs NEORSD by LC/MS/MS WSU > NEORSD
Could be a result of method flexibility modifications, standard accuracy, purity, etc
NEORSD
Method 701.0/546 vs. 544
0
2
4
6
8
10
12
14
16
18
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Method 701.0/546 vs. NEORSD and WSU Method 544
Microcystin701.0/546
NEORSD544
WSU544
MC concen
tration (ppb
)
Sample ID
Standard AccuracyCompany Name MC StandardAPExBIO MC-LRBeagle BioProducts MC-LR, RR, YR, LACayman Chemicals MC-LR, LA, RR
Cyano BioTech GmbH MC-LR, RR, YR, LW, LF, [D-Asp3, E-Dhb7]MC-RR
National Resource Council Canada MC-LR, RR
Sigma-AldrichMC-YR, RR, LA, LW [Dha7]MC-LR, [D-Asp3, E-Dhb7]MC-RR
Known issues: evaporation degradation methylation Non‐uniform dissolution
Standardize MC concentrations using UV238
Found discrepancies between labelled and measured concentrations, between lots
Could result in over/underestimation of congener concentrations
20%35%50%65%80%95%
110%125%140%155%170%
2211
406
2211
406
2211
406
2211
406
2211
406
2211
406
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
2211
407
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
411
7251
412
7251
412
7251
412
7251
412
7251
412
7251
412
L221
407
L221
407
L299
94L299
94L299
94L299
94L303
54
Standard Analysis MC‐RR
% Recov
ery of M
C Stan
dard
Lot number
Standard Purity
Enzo Standard Lot Number Observed Impurities % Impurities MC-LR 02211428 DAsp3MC--RR 1%MC-LW L30332MC-YR L30158 MC-HtyR 2%
DAsp3MC-RR L30311
MC-HtyR L30287 MC-YR, MC-LR, MC-D-Asp3MC-LR
14%
Nodularin 0614420MC-LY L30309
DAsp3MC-LR L30336MC-LA L30307MC-RR 0221407 MC-D-Asp3-RR 2%MC-LF L30373 MC-LW 2%
MC-HilR L30372MC-WR L30310
Enzo Life Science standard purity summary
ELISA Cross‐Reactivity Individual congeners run via
ELISA; measured in MC‐LR equivalents
Congeners potentially overestimated/ underestimated relative to MC‐LR; differed from kit correction factors (Issue with cross‐reactivity? Standard accuracy?)
0.6 ppb MC‐LA, 1.5 ppb MC‐RR reported as 1 ppb MC‐LR
